(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Exploring the effects of Medicare PPS on Medicaid"

Department of Health and Human Services 
Health Care Financing Administration 
Office of Research and Demonstrations 



Medicaid Program Evaluation 
Working Paper 

MPE 8.1 December, 1985 



EXPLORING THE EFFECTS OF 
MEDICARE PPS ON MEDICAID 

John Trutko 

Clare Feinson 

Larry Beyna 



James Bell Associates, Inc. 



Project Officer: Gerald Adler 



This Working Paper was developed under HCFA Contract No. 
500-83-0058 by James Bell Associates, Inc., Suite 826, 
1700 North Moore Street, Arlington, Virginia 22209. 
(703) 528-3230. 



LIST OF EXHIBITS 



1.1. Potential "Spillover" Effects of Medicare 2a 
DRG on Medicaid Program Expenditures 

2.1. Prospective Payment in State Medicaid 9a 
Programs — Changes Over Time 

2.2. Current Prospective Payment Systems: I5a 
33 States and One Territory 

2.3. Detailed Descriptions of Ten State isa 
PPS Initiatives 

2.4. Peer Grouping Methodology for Hospital 49a 
X, Pennsylvania 

2.5. Discharge Discount System 52a 

3.1. Summary of Potential Medicaid Consequences SSa 
of Hospital Response to Medicare DRG 

3.2. Factors Likely to Affect Hospital 59a 
Adaptation to DRG 

3.3. Possible Effect of Reduced Length of Stay 62a 
and Other Efficiency Measures 

3.4. The Potentially Negative Effects of 70a 
Cost-Shifting 

3.5. Negative Effects of Patient Dumping 7ia 

3.6. Mixed Potential Effects of Hospital 72a 
Strategies to Change Case Mix 



CONTENTS 



1. Introduction and Suinraary of Findings l 

2. Status of Prospective Payment 7 
Systems in State Medicaid Programs 

2.1. As of October 1984, Two-Thirds of the 7 
State Medicaid Programs Used PPS for 
In-Patient Hospital Care 

2.2. Medicare DRG Has Apparently Had 9 
Some Influence on State Medicaid ' 

PPS, But Recent Legislated Changes 
and General Cost Containment Pressure 
Appear to have been More Powerful 
Influences 



2.3. Per Diem Remains the Predominant 
PPS Unit of Payment 



14 



2.4. Synopsis of State PPS 15 

2.5. Overview of Ten State Medicaid i? 
PPS Initiatives 



2.6. State Descriptions 27 

California 27 

Illinois 30 

Maryland 33 

Massachusetts 38 

Michigan 41 

Ohio 43 

Oklahoma 45 

Pennsylvania 48 

Utah 51 

Washington 52 

Docximents Reviewed for Detailed 55 
Description of Ten State Medicaid PPS 

The Potential Medicaid Consequences of 57 
Hospital Response to Medicare DRG 

3.1. Medicare DRG Is Expected to Affect 58 
Medicaid, and Impacts Are Likely to be 
Mixed 



CONTENTS, CONTINUED 



3.2. Medicaid Costs Are Expected to Decrease 61 
from Reduced Length of Stay and Increased 
Efficiency, but at the Possible Risk of 
Negative Patient Outcomes 

3.3. Medicaid May Suffer Cost Increases and 70 
Other Adverse Impacts Due to Cost Shifting 

and "Patient Dvimping" 

3.4. Adoption of Other Case Mix Strategies by 72 
Hospitals May Lead to Mixed Results for 
Medicaid 



zrr 



CHAPTER 1 
INTRODUCTION AND SUMMARY OF FINDINGS 

In October, 1983, the Health Care Financing Administration 
(HCFA) introduced a new prospective payment system into the 
Medicare program — Diagnosis Related Groups (DRG) — under which 
hospitals are reimbursed a predetermined amount per patient 
discharged for medical services provided under Part A (hospital 
insurance) of the program. Most analysts agree that this change 
in Medicare reimbursement represents a significant shift in na- 
tional policy — one which is likely to extend well beyond Medi- 
care. For example, some analysts view DRG prospective payment as 
a force that will lead to a major restructuring of the hospital 
industry and the way in which health care is paid for both pub- 
licly and privately — affecting virtually all aspects of medical 
services in this country. It is not surprising, then, that 
prospective payment (and, in particular, DRG) has become a cen- 
tral focus in health-policy and cost-containment discussions. 

Medicare is one of two major programs that provide federally 
financed health care to large segments of the Nation's popula- 
tion. Medicaid is the other. There are major differences 
between the two programs— for example, the Medicare program is 
uniformly administered nationwide by the federal government while 
the state-administered Medicaid program varies in important ways 
from state to state. In spite of their differences, the two 
programs have had considerable influence on each other's policies 
and operations during the past two decades of growth and change. 
Given the close relationship between the Medicare and 



Medicaid programs, the recent implementation of DRG in Medicare 
now prompts analysts to ask: Will state Medicaid programs gravi- 
tate toward prospective payment, perhaps using the DRG model or 
some variation of it? How will Medicare ' s adoption of a prospec- 
tive payment system otherwise affect state Medicaid programs? 
Will the effects of Medicare DRG "spillover" onto the Medicaid 
program — for example, in shifting of costs between Medicare and 
Medicaid, or in terras of changes in State Medicaid costs and 
administrative practices, changes in levels or types of services 
provided to patients within the hospital setting, or changes in 
the way in which length of stay (LOS) is monitored for Medicaid 
patients, etc? 

The purpose of the study described in this report was to 
explore these and other questions about the possible effects of 
the Medicare prospective payment system on the Medicaid program. 
In late 1984, we reviewed available documents and held discus- 
sions with state and federal Medicaid officials, federal Medicare 
officials, and providers and representatives of provider organi- 
zations. We sought hard information where answers were available 
and informed speculation about questions for which it was too 
early to tell for sure. Our aim was to identify the most likely 
n'ajor effects of Medicare DRG on the Medicaid program. 

Exhibit 1.1 outlines some possible ways in which the 
Medicare DRG program might be expected to influence Medicaid 
program policies, practices, and cost outcomes. We emphasize 
"might" here because it is still too early to make a definitive 
assessment and generalize about actual state and hospital 
behavior and the implications of that behavior for changes in 



EXHIBIT 1.1: POTENTIAL 'SPILLOVER- EFFECTS OF MEDICARE DRG ON MEDICAID 
PROGRAM EXPENDITURES 






Medicare 
DRG 



Hospitals serving Medicare popu- 
lation adopt one or more of the 
following adaptive strategies tliat 
can affect Medicaid: 


(-) 


Potential 


• Reduce average length of stay for 
alLpaUents, including Medicaid. 


Effects 


(-) 


on 


•Promote more use of outpatient 
services for all patients. 


Medicaid 


► 

(-) 


•Promote less use of ancillary services 
for 91II patients. 


Expenditures 


— >■ 
(♦) 




» Promote less use of ancillary services 
for Medicare patients, but more for 
others, including Medicaid. 




(0 




•Transfer or screen out unprofitable 
patients, shifting them to public 
hospitals (and Medicaid rolls). 




► 

(?) 


(+) « Increase 


•Increase admissions. 


(-) « Decrease 


(?) ^ 


•Specialize in profitable services. 




(?) 


(?) = Uncertain 


♦ Improve record /billing systems. 






(-) 








States consider adopting 
prospective payment in 




Uieir Medicaid programs. 


— ^ 





volume and types of services reimbursed by Medicaid. The net 
cost changes to the program are far from clear at this point. 
Much of what is reported here is informed speculation, but we 
expect that soon it will be possible to assess at least the 
short-term effects of DRG spillover onto Medicaid. 

As suggested in Exhibit 1.1, DRG "spillover" can be 
described as influencing Medicaid expenditures through two 
sequences of events: (A) through the adaptive behaviors of 
hospitals implementing Medicare DRG; and (B) through state-ini- 
tiated adoption of Prospective Payment Systems (PPS) for Medicaid 
in-patient hospital care. At the state Medicaid policy level, 
prospective payment approaches are beginning to affect directly 
how some states determine payment for services. In a few states, 
the use of prospective payment in Medicaid has preceded the 
introduction of Medicare's DRG; but in many others, DRG is clear- 
ly catalyzing state attention and action regarding prospective 
payment. * '. ■ . ^ ' . •. .! - .. 

Some state Medicaid Program officials, aware of the cost- 
containment incentives under prospective payment, have already 
implemented some form of prospective payment, or were beginning 
in late 1984 to examine possible cost/benefits, as well as the 
feasibility, of implementing DRG-based reimbursement systems. 
For example, as of October 1984 (the cut-off point for data 
collection in this study). New Jersey, Pennsylvania, Utah, Ohio, 
and Washington were actively committed to using a DRG-based 
financial reimbursement system within Medicaid. There is addi- 
tional evidence that numerous other states — among them Minneso- 



ta, Michigan, and North Carolina — were considering DRG prospec- 
tive payment as a serious alternative to their current systems 
for financing in-patient hospital services. 

The recent state trend toward PPS to reimburse Medicaid 
costs is described in detail in Chapter 2. Those findings can be 
briefly summarized as follows: 

• As of October 1984, 33 states and one territory were 
using some form of PPS to reimburse for inpatient hospi- 
tal services under Medicaid. Eighteen states and three 
territories were still using retrospective payment based 
on "reasonable cost." 

• The adoption of PPS has increased rapidly over the past 
several years. while Medicare's implementation of DRG 
has been influential in this development among the state 
Medicaid programs, it appears that general pressure to 
contain costs and administrative flexibility provided by 
recent federal legislation were more powerful influences 
on the state Medicaid trend toward PPS. 

• As of October 1984, the most common form of PPS was the 
per diem model (16 Medicaid programs). Four programs 
were using a DRG model, and the remaining 13 were using 
per admission, per case, per discharge, budget review. 
etc . 

• For the most part, state Medicaid programs have been 
eclectic in the development of their PPS and have made 
many adaptations of basic models to fit unique state 
circumstances. No two systems are identical. Even among 
the four DRG systems, there are important differences in 
such system characteristics as classification scheme, 
weighting procedures, degree of risk sharing, etc. 

At the hospital level — the other major area in which 
Medicare DRG can be expected to affect the Medicaid program — 
DRG could have positive or negative effects. For example, to the 
extent that hospitals are able and willing to extend new DRG- 
inspired efficiencies in serving Medicare patients to all pa- 
tients with similar diagnoses and service needs, Medicaid will 
benefit in the area of reduced costs. However, to the extent 
that quality of care suffers for the sake of efficiency and new 

4 



care patterns are generalized to all patients, Medicaid reci- 
pients will suffer too. On the one hand, Medicaid may generally 
benefit from whatever improvements in administrative practices 
that hospitals make (e.g., record keeping, billing, etc.) in 
order to meet DRG requirements. On the other hand, Medicaid may 
incur increases in payments for costly patients and services 
shifted away from the Medicare system, and so on. The dynamics 
of possible hospital behavior in response to DRG is very compli- 
cated and as yet unclear; as such their potential effects on 
Medicaid services and costs are uncertain. 

This topic is explored in detail in Chapter 3. Findings 
with respect to the expected effects of DRG at the hospital level 
can be summarized as follows: 

• Hospitals could respond to Medicare DRG in five ways: 
(1) devoting increased attention to reducing length-of- 
stay for all patients, not just Medicare patients; (2) 
improving the efficiency of service delivery for all 
patients; (3) shifting Medicare costs to other patients, 
including Medicaid patients; (4) "dumping" financially 
undesirable patients — Medicaid and Medicare — into 
the care of public hospitals ; and ( 5 ) manipulating case 

li-iL^.' : mixes to provide a higher return on costs. 

• Hospitals' need to change policies and practices in 
response to DRG is likely to vary with the relative size 
of their Medicare caseload and the extent to which they 
were already operating efficiently prior to Medicare 

• Hospitals' ability to adapt successfully to Medicare DRG 
is likely to be influenced by several factors, includ- 
ing: private vs. public ownership, stringency of 
Medicaid reimbursement policies in their state, opportu- 
nity to specialize and collaborate with other hospitals, 
ability to develop long-term care alternatives to hospi- 
talization, ability to shift loss-producing patients to 
other hospitals, ability to influence physicians' admit- 
ting and treatment practices, etc. 

. • Medicaid costs could be expected to decrease as a result 
of DRG-induced reduced length of stay and increased ef- 



ficiency, but patient outcomes could suffer. Possible 
negative outcomes for patients include increased read- 
missions due to premature discharge, increases in incor- 
rect or delayed diagnoses due to curtailed use of tests 
and other diagnostic procedures, and increases in 
complications due to reduced attention per case by medi- 
cal personnel. 

Shifting of Medicare patient costs onto other patients, 
including Medicaid patients, and "dumping" unprofitable 
Medicare cases into public hospitals would increase Med- 
caid costs. 



CHAPTER 2 

STATUS OF PROSPECTIVE PAYMENT SYSTEMS 
IN STATE MEDICAID PROGRAMS 

This chapter contains the results of an exploratory review 
of the 54 state and territorial Medicaid programs in order to 
ascertain how many use prospective payment systems (PPS) for 
Medicaid in-patient hospital reimbursement, and what types are 
currently in place. The review included discussions with offi- 
cials in all ten HCFA regions, interviews with many state Medi- 
caid officials, and review of recent reports, documents, and 
other secondary sources. The objective was to identify all the 
states with a Medicaid-involved PPS in place as of October 1984, 
and to develop brief descriptions of each system for comparative 
purposes. To capture the complexity and diversity of individual 
systems, detailed descriptions of ten individual states were then 
developed. ' 

The first section is a broad overview of the use of PPS 
throughout the nation. Changes over time and the influence of . 
Medicare Diagnosis Related Groups (DRG) on PPS development are 
discussed. In Section 2.2, detailed descriptions of ten state 
PPS approaches are provided, following a comparative review of 
several key characteristics of those systems. 

2.1 |s of October 1984, Two-Thirds of the State Medicaid 
Programs Used PPS for In-Pati^5t itospltal ci?i" ^^^^ 

As of October, 1984, 33 states and one territory were using 
some form of prospective payment. Eighteen states and three 
territories continued to use the Medicare "reasonable cost" re- 
trospective payment approach for Medicaid- funded in-hospital 



care. One state was using a prospective rate setting system, but 
at the end of the year routinely reimbursed hospitals whose total 
costs exceeded the rate amounts. 

For the purposes of this report, a state PPS had to include 
Medicaid in-patient hospital reimbursement to be counted in the 
PPS colximn. (South Carolina, which had a PPS, but only for 
nursing home services, and West Virginia and Wisconsin, whose PPS 
did not include Medicaid, were therefore considered non-PPS 
states.) Further, a PPS was defined as having the three charac- 
teristics outlined in HCFA/ORD's October, 1982, "Report on Hospi- 
tal Prospective Payment Systems Mandated by Section 2173 of 
Public Law 97-35." According to that report, PPS provides hospi- 
tals with an explicit set of payment rates (per service, per 
diem, or per case) or, in budget review systems, with implicit 
rates for the same units, that: (1) are determined in advance 
and fixed for the fiscal period in which they apply; (21 are not 
automatically determined for any individual hospital by the level 
or pattern of its present or past incurred costs or charges, and 
(3) are generally intended as payment-in-full for the specified 
unit of service. 

The most problematic yet the most important of these fea- 
tures is the third. Systems vary widely in terms of how strin- 
gently hospitals are held to the prospective rates when actual 
costs exceed them. In many systems, adjustments are made at the 
end of the rate year and hospitals are reimbursed for additional 
costs, but this is usually on a partial and case-by-case basis. 
Automatic and full adjustment systems were not counted as PPS in 



8 



this review. 

The HCFA report also defined PPS as allowing hospitals to 
]ceep all or some portion of the difference between the rate and 
actual costs (when the cost is lower). A few state programs that 
require hospitals to reimburse the state if actual costs are 
lower than the prospectively-set flat rate ~ Georgia, for exam- 
ple — were retained in this current count of Medicaid PPS. 

Preliminary efforts to apply precise definitions and cate- 
gorize the various forms of PPS have been difficult, not only 
because change is very common among the 54 systems, but also 
because almost every system is a unique combination of features 
for which there are no standard definitions or common use of 
terms. 

The adoption of prospective payment has increased quite 
rapidly over the past few years, as indicated in Exhibit 2.1. 
According to the October, 1982, HCFA report, only 14 of the 50 
states had PPS in operation as of April of that year. Fourteen 
months later, the number had climbed to 20, and then 16 months 

later to 33. 

'-■ ' 

2.2 Medicare DRG Has Apparently Had Some Influence on state 
Medicaid PPS, But Recent Legislated Changes and~Generai 
Cost Containment Pressure Appear to Have Been More Powerful 
Influences 

While there has been a surge of PPS adoption coincident with 
the implementation of Medicare DRG, it appeared from our 
discussions with state and regional officials that the states' 
PPS activity has been primarily a response to general cost con- 
tainment pressures, and the increased ability of the States to 
modify their systems as provided for in recent federal legisla- 



Exhibit 2.1: Prospective Payment in State Medicaid Programs 

Changes Over Time 



Apri 


1 '821 June 'SS^ 


====================== 




PPS 


14 . 20 


Per Diem 


6 , 11 

-* 


DRG 


— 


Per Admission 


1 


Per Case 


1 1 


Per Discharge 


1 1 


Rate of 


% 6 


Increase Control ' 





October '84^ 



33 

16 
4 
2 
1 
1 
1 



Budget Review 
and Other 



8 



Retrospective 
Total 



36 
SO 



30 
SO 



17 
SO 



X: "Report on Hospital Prospective Payment Systems Mandated by 

Section 2173 of Public Law 97-35," DHHS/HCFA/ORD, October, 1982. 

2 

Tom Fulda, "Medicaid Hospital Reimbursement Fact Sheet," 
ORP/DARS, circa October, 1983. 



Discussions, document review: September-October, 1984, 



/:j^' 



9a 



tion. The PPS movement may have been facilitated by the federal 
lead, but it existed before the recent changes in Medicare. 
Several states had already adopted a PPS or were considering it 
long before DRG was begun in Medicare in late 1983. Some have 
been using PPS approaches since the early and mid-1970s. 

All federal and regional administrators with whom we spoke 
agreed that the prime motivating force behind the trend toward 
PPS has been cost containment, an attempt to maintain in-hospital 
care costs at a level which the states can still afford. often 
the cost containment issue is closely linked to state budgetary 
problems. if state governments have a significant budgetary 
problem - that is, they are "fiscally strapped" in the words of 
one administrator ~ then they are more likely to turn to PPS as 
an alternative for reducing Medicaid budget increases. One 
regional administrator noted that PPS has given states the 
ability to "reshuffle the deck," to reduce Medicaid costs — 
where it would be impossible to introduce significant budgetary 
changes under the existing payment system for in-patient hospital 
services. 

State efforts to contain Medicaid costs have been greatly 
facilitated by increased flexibility in reimbursement policy 
provided by recent federal legislation. Changes undertaken by 
Congress in the past three years have had two general themes. 
First, Congress has moved to increase substantially the states' 
flexibility in determining the structure and policies of their 
Medicaid programs with the expectation that they will use this 
flexibility to help control costs. Second, Congress has mandated 



10 



several direct changes to the program, including, a reduction in 
the rate of Federal financial participation, to slow the cost 
spiral. Although some of the recent changes constrain state 
options and add to program costs, the overall emphasis of recent 
federal legislation has been on state flexibility and cost con- 
tainment. '■'' " ^^- 

The three legislative initiatives that have had the most 
influence on the present " shape of the Medicaid program are the 
Omnibus Budget Reconciliation Act of 1980 (OBRA 1980), the 
Omnibus Budget Reconciliation Act of 1981 (OBRA 1981), and the 
Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA). 

The Omnibus Budget Reconciliation Act of 1980 (P.L. 96-499) 
introduced the first round of major Medicaid changes. This 
legislation replaced "reasonable cost" as the basis for reimbur- 
sing Intermediate Care Facilities (ICF) and Skilled Nursing Faci- 
lities (SNF) with reimbursing according to rates which are "rea- 
sonable and adequate to meet the costs which must be incurred by 
efficiently and economically operated facilities." Other provi- 
sions included safeguards to discourage transfer of assets by 
hospitals and allowing "swing-bed" arrangements in rural hospi- 
tals. 

The passage of the Omnibus Budget Reconciliation Act of 1981 
(P.L. 97-35) strongly demonstrated the federal government's 
concern for the rapid escalation in government expenditures for 
medical care for the elderly and poor. Section 2161 required 
reduction of federal financial responsibility under Title XIX 
programs (Medicaid). In addition, to compensate for this 
reduction in financial responsibility, OBRA extended greater 

11 



e 



flexibility to states in methods for making Title XIX payments to 
hospitals. 

Section 2173 of the 1981 Omnibus Budget Reconciliation Act 
deleted requirements that reimbursement for in-patient hospital 
services be on a reasonable cost basis. Under that law, hospital 
services under Medicaid were to be reimbursed according to rates 
that the state assures are reasonable and adequate to meet costs 
that efficiently and economically operated facilities must incur 
to provide care in conformity with applicable State and Federal 
requirements and quality and safety standards. The statute also 
required that methods and standards used to determine payment 
rates must take into account the situation of hospitals serving a 
disproportionate number of low-income patients with special needs 
and provide that reimbursement for patients receiving an inappro- 
priate level of care will be made at lower rates. Payment rates 
must also be adequate to assure that recipients have reasonable 
access to care of adequate quality. 

States must provide for uniform cost reporting and periodic 
audits. The regulations (42 CFR 447.255) require States to 
submit other related information including estimated average 
payment rates and the impact of these rates on availability of 
services, type of care furnished and extent of provider partici- 
pation. Finally, the regulations (42 CFR 474.272) establish 
upper limits on payment rates for in-patient hospital services 
under Medicaid which are no higher than would be paid for such 
services using Medicare principles of reimbursement. 
- TEFRA's effects on the Medicaid program, while less exten- 



12 



sive, were still important and continued to emphasize cost con- 
tainment. For example, TEFRA introduced "nominal" co-payments on 
certain mandatory services for specified groups of the categori- 
cally needy. Co-payments were prohibited, however, in certain 
important areas, such as emergency services, family planning 
services, and pregnancy-related services. 

Although the growing need to contain costs and the increased 
flexibility provided by recent federal legislation appear to have 
been the two most powerful influences on states ' movement toward 
Medicaid PPS , Medicare ' s introduction of PPS is also viewed as a 
motivating factor. According to one administrator. Medicare DRG 
provides an established methodology. Another regional official 
noted that states are likely to follow the line of least 
resistance. For example, in at least one state a major factor in 
the selection of a DRG-based system was that, since the hospitals 
are already adjusting to the federal DRG system, the adoption by 
the state of a similar system would mean that hospitals would 
have to make only one major adjustment instead of two. 

Although states are generally unlikely to adopt prospective 
payment merely in order to be in line with Medicare, the DRG 
system has created a clear precedent and model that prompts 
consideration. Many states have been willing at the minimum to 
invest resources in the form of a commission or task force, to do 
one or more of the following: (1) monitor effectiveness of 
Medicare DRG, (2) examine the possible application and implica- 
tions of Medicare DRG on the state, and (3) begin drawing up 
contingency plans for establishing PPS. In fact, officials in 
some states indicated that they were waiting for demonstrative 

13 



results from Medicare DRG before making a decision about PPS. 

In October, 1984, four states were using DRG systems. They 
are New Jersey, Ohio, Utah, and Pennsylvania .( 1 ) Washington 
planned to implement DRG in November, 1984, and Michigan, Oregon, 
and Connecticut were planning to adopt it in 1985. Several other 
states were considering DRG but were taking a "wait and see" 
attitude for the time being (e.g.. North Dakota, South Dakota, 
Nebraska and New Mexico ) . 

2.3 Per Diem Remains the Predominant PPS Unit of Payment 

Exhibit 2.1 also details at a gross level the types of PPS 
in use over time. The dominant type of PPS in use in 1982, 1983, 



(1) "Diagnosis -Related Groups" (DRG) is a system of 
classification of medical diagnoses on the basis of clinical 
consistency and similar resource use, developed at Yale 
University for use in reform of Medicare expenditure. This 
classification system is not in itself a method for determining 
prices, but rather a tool to be used in a pricing system. For 
the purposes of this paper, our definition of a DRG-based 
prospective payment system includes the following factors: 

— use of the Yale University case classification system; 

~ use of a relative value scale to relate resource 
consumption of the individual DRG classifications to each 
other ; 

— use of a case-mix index to reflect the relative frequency 
of discharges among hospitals; and 

— use of these factors to determine pricing, along with an 
adjusted average price or cost per case. 

This definition includes four states: New Jersey, Ohio, 
Pennsylvania, and Utah; even though their pricing systems are all 
very different, they share these common factors. It does not 
include states such as Maryland, which, although it uses the DRG 
classifications for setting utilization limits, does not use DRG 
for determining the rate of payment. - -• :^ 



14 



and 1984 was the per diem approach, which limits hospitals to a 
raciximvim charge per day of service provided. In late 1984, .at 
least 16 (or almost half) of the 33 state PPS programs involved 
per diem as the unit of prospective payment, while only four had 
adopted a DRG model. The others appeared to be divided among per 
diem, budget review, and a variety of other approaches. These 
are far from exact or mutually exclusive categories since many 
states have been very eclectic in the development of their sys- 
tems. We used these gross labels primarily in order to compare 
this update with the 1982 and 1983 descriptions of previous 
reviewers. Per diem was the basic unit of payment in approxi- 
mately half the existing PPS programs in those two years as 
well.(l) 

Some per diem systems include different rates for different 
types of services; some include length of stay limits. While a 
large nxanber of states are using a per diem unit of prospective 
payment, this approach may give ground in the future to the DRG 
method, once that approach has had time to demonstrate its effec- 
tiveness in containing costs. 

2.4 Synopsis of State PPS. Exhibit 2.2 presents a very brief 
description of the PPS in each of 33 state and one territorial 
Medicaid programs as of October, 1984. The exhibit groups the 



(1) Several PPS that at first were labeled rate of increase 
control were later placed in the per diem or another of the 
categories, because they involved the use of a specific type of 
unit of payment. For this reason, the apparent change over time 
in the number of Rate of Increase Control systems from five to 
six to one more likely reflects a change in descriptions applied 
from one review to the next. 



15 



Slal* 

■■■■■■«■•«! 

ONO 



Uali of 
PcyiMal 



oaa--ptr 

illtehirci 



™i«iT I.I. amiiocT PHrwPBCT.vB rAW«.r av5T«>«. „ „*tm am, , TK.HHm,«y 
" •— 

IfdMi OaierlplUa 



..............a.... 

lae|yil«<l lUk Sk«rlii( 

■*■•"..•."■.....«......•.»..,.,, 



....... *.....«.a..ss....x. 

Cil«al af Pvlvr* PUai 



by ComaU.iaaar af H..lik. Th. pratp.ell." 
ciaa-iali-batad ralmburi*««ni atlabllthat a 
Iliad pavmaal rata par lyp, ,f „,,. „,,, 

•'•.""iSi'*'""'""'* ••< "..ad aa r«0. Sip- 
arala DRO prlc, „, „, ,„ ,„^ h«,p|,,| 

batad ea lit ual<|ua aharaelar Itl lea and 
•paairic aaila. 



All payari 



Provldar abiarbi 
diffaraaca bal- 
wata rait and 
actual eatla. 



PPI kai iloaad In- 
pallaal aipandllura 
Incraattt la II. |« 
la lilt and 14. }« 
la nil. 



Ohia 



0«0--paf 
dltcharc* 



Ul 



Panaayl 
*anla 



Oaa--par 
dlicharga 



aara claiiiricatlaa ty.i.. ,„4 Ohl«-,p,c| lie 
-alfbla and paar ffauplag batad on .a,« ra at 
r.rllalp.i a. .,d aan^lla... .,. ^..dil.M 
(Saa Eaklbll |.| |ar mra dalall.) 



Tha llala Vadlaald a(aaay a*iiiaii|.^i'ika 

lytlam .bleh utat Madlaara nno clait illatllont 
•>4 • al.-.pa.lll. «al,h|.. a aon^ a, pIIJ "' 

f"u!" i^:"^• ^''•" " •"•"« •••luaiio. 

(Saa Eihlbll I.J lor mara dalall.) 



Madlctid 
aniy 



Pravldar abterbi 
diriaranca bc- 
Ixaaa rata and 
actual coitt. 



Medicaid 
aalv 



(ama at aba»a. 



Tba lyttam It bain* 
pbatad In avtr 
thraa yaart (Ifit- 
IIKl. Evanlually 
priaat m| I | ba la- 
tally batad an paar 
■roup avarataa (and 
■at at all an hot- 
Pilal-tpacllle 
ratal). 



Phata-ln It ilmllar 
to Obla. 



Ulak 



OHO- -par 
dlieharf a 



nmi-bii.d lyitM. .kich ittwl'Vtt',n',\l','ll''u,,l'i,' 

ela.tlllc.il., tyttam and ralallva •.cl.hli. 
but utinc ilata cl.lmt data ta r.duca WtO 
payment ratal. 
(Saa Eihlbit t.) for more dalall ) 



Medicaid 
only 



II aelual eaiti 
are leit tban 
W" lar«et rale, 
ttaie payt cottt 
plui lOtk luo la 
laract rate)i ir 
aelual eettt ei- 
cead lar(al rale 
tban tiate payt 
larRel rata un- 
til cottt are 
*»%, above tarnel 
rale, Ibeo «0« 
ol additional 



Tha Hate capecti 
te eventually de- 
velop a atata- 
tpeclllc prieint 
lyitem. It I, .1,0 
cianilnln( the (eat- 
abl Illy of movlnc 
In a lelaetlve pro- 
vider conlraetint 
proiram. 



EXHIBIT 1.1. CURJJEWT PRnapiiCmVI PAYMRKT SYSTCMSi II STATES AND I TEHIIITUaY CDKrimiED 



Per DIcn 

AlabftiM 



Ptr diem Slele ((eney eelt hoapl I* I -ipeel ( U per diem Mcdieeld 
relet eeleuleled on beie-yeer eoili trended only 
loroerd by Inflillon indee. 



Hoipliela ebeorb 
difrerenee bel- 
xeen per diem 
rele end eeluel 
eoti t , 



Cellfornle Per diem Seleellve Provider Contreelinf Profrem (SPCP», 

under whick Ihe tlele ne(ollelet eonlreelt 
Indlvlduelly xllh eligible provider! I* lleellh 
reelllllei Plennlnn Areet (MPPAil. Rural ereee 
(II e( Ihe 1)1 IIFPAi) remeln on e fee-for- 
•ervlee lyilemMllh tlrln(enl coil levlnn 
IneenI I vet . 
(See Eihlbll I.) for more delell.l 



Medlceld 
only 



Provider! ebiorb 
dirferenee bel- 
oeen eonlrecled 
per diem rele 
end eeluel 
coil t . 



The eurreni eon- 
Ireclini tytlem 
covert (boul 10% ot 
Medl-Cel budfel lor 
In-pellenl cere. 
The title It devel- 
oping t pedltlrlct 
ditgnotli cleiilM- 
etlloit Ihel It 
eipeeled lo enable 
ehlldren'i hotplltlt 
la be a perl of Ihe 
protpeallve tytlem. 



Colorado 



Per diem 



Slale Medlceld egeney aeli hotpllel-tpecl f le 
pretpecllve per diem relet for ell hotpllalt. 
Baae-yeer eoalt tre trended foroard utlng an 
Inllellon (actor. LOS and reatont (or admlt- 
iloji are tubjeel to review. No relroeellve 
ediutlmenit ere medei' Intent I ly of tervlee, 
new tervleet, etc. ere contldered during rete- 
lel t Ing procett . 



Medicaid 
only 



Hoepltala abaerb 
loitet above end 
relein tevlngi 
below per diem 
relet . 



"""Ol* f*' <'lem The epproprlaled ttele budget It divided by 

prelected bed/dey need to yield a per diem 
*l<*l will cover pro|ecled need without eieeed- 
Ing the ttale budget. 
(See Eihlbll t.l (or more detail.) 



Medlceld 
only 



llotpltala ebtorb 
difference be- 
tween per diem 
rele and eeluel 
cotit . 



Stale It eontlder- 
Ing SPCP progrem 
timller lo 
Cellfornle. 



Per 
diem 



Provldert ere held lo their prevlout yeer 
cotIt plut a tmell Increata whieh It calcu- 
lated by applying verloua merkel betkel 
Indeiet. Provldert are given their rele of 
relmburtcmenl ei their fiteel yeer beglnt. 



Medlceld 
only 



Provldert ebtorb 
difference be- 
tween per diem 
rele and actual 
eotla. 



EXHIBIT i.ii cuiuiEKr mospBcrivB pavmekt SYSTvai ii statu and i mwiTQiiv ODKriimeo 



KtHKI 



P*r 

dim 



Similar lo loo* 



Medicaid 
only 



frovldara abaorb 
dlffaraiiaa ba- 
l«aaa par diam 
rata and aelual 
eoala. 



Kanluaky Par dIam Tha (lata afaney admlalalara a ayilam la «Mali 

uppar llatlll ara plaoad on ralmburtabla colli. 
Tka limit la IIM of lh« Malfhlad madlan par 
dIam aoila for aach of four hotplial paar 
|rouplii(a. Contpllanea la mandatory. 



Madlcald 
only 



Provldara absorb 
diftaranaa bal- 
maaa par dlaoi 
rata and actual 
eoala. 



Maryland Par dIam/ Proapaatlva rala-aallUf and bud(at ravlaa All payar lioapltala abaorb As of Oat. 1(11, 

fliad par ayitam xhiek ailabllahai par dIam ralai/fliad Ika ditfaranca tha atala baa addad 

unll of par unit of larvlca ralai In ravanua-produein* balwaan Ika aal- a IMtO-baaad alamanl 

aarvlea eanlara for aack koapltal by comparlni dirael abllakad rata to tka rata ravlaa 

rataa aoala wllkin paar irouplnfa. and aelMal proaaaa, undar 

(Spa Eiklblt 1.1 for mora detail. I coata. Mklek Ika atate 

paya Itm of Ike 
a«era(a Maryland 
Madlcald W» oltkln 
a fivan Dun 
elaasi float Ion. 



MlaalsilppI Per diem Proapeellva par diem rale eel for provldera Medicaid 

raaullinf frem audited baaa-year coals only 

trended lorniard by inflation. 



difference be- 
laecn per diem 
rata and ectua 
coats. 



Mil 



Par diem 



Similar to Iowa 



Medieeld 
only 



Prevldera ebso 
difference be- 
teiecn per diem 
rale and aclua 
coals. 



Par 
dien 



Similar to Iowa 



Madlaald 
only 



Provldara abao 
difterenee be- 
tween per diem 
rale end aclua 
coats. 



Tlia state la very 
Interested in DM], 
auck as tka Ohio 
modal. Tkare kava 
been no decisions 
yet, and OftO oould 
not be loiplemented 
before July. lilt. 



\ 



EXHIBIT t.ti cuiuiRKr PRoaPKn-ivi PAYuurr sYsroai ii statu and i TEiuiiit)iir oontihueo 

M.» York P«r dl<m Th. D«p.rlin.i.l or H.*llh Coomliilonor e.rll- All poyori HotplUI. .biorb 

fill propoaod hotpllol-ipteUU p«r dUm roUt taooadv* «ailt 

for Modleold to Ik* DIro.ler of Iht Dudxtl. ,„d ktop •"!»■• 

Thojt rdoa (lelol ln-p*ll«iil cotli divldod by ik.i roiull Ireni 

p.ll.nl d.y.l .,. ....bll.h.d fo, .pi- .omini U undir 

I.U t. . fu.ronl.od r«««nu. o.p. Th., .r. ih. ,i,.»u. «tp. 

btiod on ••(u*l boo y.tr eo.la Ir.nd.d (or- 

a.rd for InfUllon and ir. adJutLd for ■"' "' "' 

*•>•"»•• I" »olum« and caaa-mla, aarvlcaa addad ■ ■ i ' 

or dalalad, and labor eoal Inaraaaaa. Allew- 

abla eoBla ara llmllad to avarafa eo.l el paar 

froup plua }«. Bad dabi and aharlly cara ara 

aovarad. Each ho.pllal haa a apaeirie l-OS 

ilandard baaad on eaia mli. 

M. Carolina Par dian. Ho.pUal-.pael rie par diam rata, arriaioii Medicaid I'loipluii'ibiorb 

ba.o yaar eo.l. Ir.nd.d forward u.ln| impA only difrar.naa ba- 

markal ba.kot. ,„,„ |h, p., 

diani rata and .. ,;.,, 

actual eo.la, 

okiahc™ ra, di.« j;:'::«;i:;, ;:j«;::iiii«"'jj«'f-^i«jj;i«b: M^iuiii ;;;;•»■;;;;; pmm;;;;;;^^!"" 

nana, lael 1 1 ly-apeel fla per diem raimburaa- only ..ila.. .k!>- ik.i 

men! ra aa. Slrln(anl ulllU.llon eonlrola tli.rihl ll^«l- 

and raelea of palleni eare. . .. . '"'?'•'»•»- 

(See Eahlbit 1.1 for mora detail.) l!!!??,*' **"'"• 

> <i.t apaeltlve ayelem, 

there kava been ao 

Inaraaaea In eoala 

for In-pallanI 

eare. 

Tenneaae. Par dIam Preapeellve per diem rale, eel on'ba.Ty^^r M^dtedd S^.i.^'»'^bove 

aoale trended tor-erd by Inflation Indai. only 

VIrilnIa Per diem/ The alale'.*H«IU*iirii«rc«I'i;i;i« Aii'ihir'oI'M;;;];;;;^;;;: 

bud.el Connlaalo. aaaaaae. •reeaonablene.C of ko.pl- baaed are midi Hel- 

r.»Ie- I.I b-d«at.. Celll.,. on operatln. eoala a?a payara I. J t.TI aia.rj 

baaad on median par diem eoat. for aeek ef ieJeral. iny aoal. abo»e 

f.rty-aa».« peer .r..pln|a pi a.elal.r p.y.kl.trle tkl rate. V,V 

of approilmalely i-« pereenl. Coal. IneUde and o»l- for Ikal y.Ir. 

patient eere, bed debt, ckerlty. edu.allon and palleni ' 

re.aarak, eepllal eoala, taiaa, and return on eurgleel 

aqulty. Hoapllala muat parllalpele In re*le» koapllala. 

and comply wllb ratea. 



■XHIBIT «.«. dlMBfT PROarKTIVI PAVHDrr aVSTMli 11 CTATCa AND I TmillVMIT OOKTIMOKD 



Qui 



Par dltoi 



Proap«allv«|y RcgalltUd par 4Imi ral* p*l4 
far In-kaapllal cara. 



■•■■•■■•■■■••■■•■■■•■■■■■■■■aaa 

Madtaald Preyldara abiorb 
aaly dltfaraaaa ba- 
iMaaa par diam 
rata aad aalual 
aaali . 



■■■•■••»••««■■ 



■Hd(al Raalt 
Alaaka 



aHdgal Tha ilala rala-»allla( aonnlialea eparatai a 
■a»la> proapaallva budgal aad rala-iallU* •ytl*i«| a 
unirera i«l of aaadllleai far paymanl aad a 
paraanltf* laaraaia ara appllad la dalarmlalac 
aach koipllal't Hadleald budgal. All kaa- 
pllali mill aamplf. Tka lytlam bafaa U 1114. 



Madlcald 
OMly 



Hoipllalt do nol 
abforh aboira- 
rala eotia If 
Ikay aaa doou- 
*aal Iha aaili. 
Tha Comal 11 loa 
makai eata-by- 
aata dalarntlaa- 
II oat. 



Tkl* la a naa ayi- 
lam, daaarlbad at a 
vary llbaral ona by 
ilala afdcUlt. 
Ratal Includa ralm- 
burtamtnl for Sar- 
vleaa aol ralm- 
buriad bafora. 
Maay lalarlm ralaa 
ara uadar appaal by 
kotpllali. 



Florida 



audcal 
■avlaw 



Hotpllal aeil-aoalalwMai beard aAalaUlart a 
bodgal aad rata ra«lai» aytlam far all payora. 
la Mhlak koipllali allb a rata of ravaaaa 
croailk ibova »t% af ill paari li lublaalad le a 
aomplala budfal ravlaw. Comparliea li aiada oa 
bull ol frail koipllal ravaaua par adlvilad 
adiBliilaa. Tkaia ralai ara valualary far 
payari alkar Ikaa Madlaald. Hadlaald ralai 
ara baiad oa klilorlcal aeili Uflalad forward. 



If rata af ai- 
paadllHra (roaik 
It la aiaaii af 
l(% •! paari la 
(reuplai, koipl- 
lal Biay ka ra- 
quirad lo rifuad 
a (l«aa amoval 
la Ilala Indl- 
laal aira ftiad. 
Provldari aiay 
•al aiaaad Madl- 
aald rata. 



La(lalalura appro- 
vad a diraallva far 
kaipllali and In- 
larari lo aalabllak 
a praipaallva pay- 
■aakl ■arrancamaal* 
Ikal laaludat 
•pillfia flaanelal 
laeanlivai lo 
cealila aotli aa of 
Oalebar, lilt. 



Malaa Budaal Tka Ualaa Haallk Cara Plaaaaa Conialiilaa aali 
tavlao/ praipaalUa llmlla aa aaak koipllal'i avarall 
Ravaaua ravaaaa. lavaaua llailli ara ailabllikad ky 
Cap aablratllac kaipllal raiaaraai from llaaaalal 
raqalramaaii aad Ikaa addlag laparala dadua- 
llaaa. Akoapllal'i flaaaalal raqulramaali 
ara dalarmlnad by adlaillng baaa yaar aipaaaaa 
by laflalloa, aaia-aiii voluma, allOKoaaai far 
faelllly/aqulptnaal aaili, and aoili of pro- 
laali approvad uadar aarllflaala of natd. 
Madlaald illoMabla coala ara lubjael la TCrRA 
eoil laaraaaaa. Pari lalpal loa In ravlax and 
aompllaaaa wllk ravaaua aapa ara mandatary. 



All payar 

(aiaapi 

Madlaara) 



Hoipllala ara 
paaalliad for 
aiaaadlag rava- 
aua aap and ara 
rawardad for 
aomlm la bala« 
aap. 



EXHIBIT t.ti cxnutDrr PRosrecTiVE rAYMEirr sysitmsi ii statu and i touiitoiiv oornmieD 



Mcitiehu- 
• •III 



Budt«l 



(•l«d lor «(eh hoiplUI wllh ••oh p^ytr r«- 
iponilbW (or • proparllo* »l th* r^v^nu* Ih^l 
corr^tpandi le Ih^ properllon of Ih^ hoipllil't 
• trvleo ui*d by III lubierlbiri. 
($•• eiklbll t.l tor mor* doUII.) 



All p^y^ra 



Hoipllili Ihil 
lldp (Olll boloM 
rov^nu^ ••p •n- 
IllUd lo ■■••p 
bal*n<^ II dli- 
•r^llontrf 
In^ooi^i If Ih^y 
•■•••d r^v^nu* 
eip Ih^y mull 
to»or Iho ihorl 
foil. 



If M«dlo*r* wilvar 
li itol rin^w^d Ih^ 
• lalatrlll n^^d lo 
iii*k^ b^il* chin(*i 
In Iho oporollon 
of Ih* current 
proip^el !¥• 



Ml^hlgm Bud(*l Preopoollvo budfol rovlow (ydira undtr whioh 

ll^«T*« p^ymiali or* band en proipoollv* oiIIimki sf 
• •eh hoopllil'i budf«l. R^h^bl 1 1 ••I Ion ond 
Inaohlnf hoiplloli ir* iicludid. 
($•• Eahlbit t.l for moro d^lill.) 



M^dleild 
only 



r^olllll^i ir^ 
not r^lmburiid 
If Ih^y fo ovor 
budfili r^eilvo 
Ineinllvo piy- 
m*nl If Ih^y ar* 
und^r . 



Mlehli^n pl^n^ lo 
ImplwKinl • DRO- 
b^i^d lyilui oarly 
In nil, which olll 
b^ **ry •Imll^r lo 
M^dle^ro nRO. 



Khodo B<id(«l till* budfd effloo, Bluo Cro^a, •nd Hoipllal 
filand RavUa Aiaoolallon of R.I. 1*1 a *mailaap,* or alala- 

Hld* limit on hoipllal aip^ndl lur*i. Th^n, al 
Ih* koapll^l ■•v^l, ih^ budi^l r^vl^a prooat^ 
feeuioi on lnar*m*nl^l ch^ng^^ from Iha cur- 
rant la Ih* proip^cllvc y^^r. InfUtlon, 
volumo ehcn(^i, ind n*« and aipandad lardaaa 
ar« eonildarad. Aflar ravlaw, aaoh hoapllal 
ailabllih*^ a ichcdul* of «h*r|*a. All non- 
f^d^ral hoipllala muil pirllclpcta and comply. 



Medicaid 
and Blu* 
Crea> 



Washlnflon Budcol Until Novomb^r IS, 1114, W^ihlnglon hoi a rat* 
R*vl*w/ ••llinf (nd budf«l r*«l*w lyilcm thil work) 
(p^r Ihroufh a rovlcM of m^|or coil onlcra vllhln 
patient (ach hoipllal. Budgat aeraana arc developed 
dliehirgel bii«d on all hoipllala acroii the atile. 
(See Eihlblt 1.1 for more delell.) 



Medicaid 
only 



Colli above the 
bud(et icrien 
ara dliallowed 
by Iha alala. 



Weahlnglon wl 1 1 Im- 
plement mn-beied 
PPS In November 
■hich uici the Medl- 
eere DRO cleii I f I - 
eetlon lyitem, 
■elghle, and hoi- 
pllal-ipeelf Ic 
average eoiti per 
ceie. Kith Medicare 
relei id|uiled 
ecreii the boerd. 



EXHIBIT l.tl OlMeNT rNMPKTIVt PAVMRKT tVaTMSl 11 STATCa AMD I TOMITORV CDKriHUED 



Olkar 



Arliont Par Tk* ilaU li ••■dualliif • lkr««-y«ar dtmon- McdUald, 
ecplU/ (Iralloa la«el«lii( prt p«|4 tiplUltd rimiiclnf •Ida 

•••Iraola Ikroufk aooipallllva conlraali le prevldarti amplayaai, 

prIiMry «ara pkyalalana aarva at Ralakaapari loma 

undar llallad eaaauinar traadom-of-cholca. oikari. 

Cokiumari akara aoala. A privala eonlrael 

admlnlalralar makafaa Ika ayalam. 

Oaarcit Par aaaa *lala •(«aay aala proapaallva par cata ralaa Matflcald Haapllala mual 

baaad ok klilorlaal avarac* coala par only abasrb aoala 

admlaalen. abova rial rata. 

Tbay mual ralm- 
buraa alala II 
Ikair aalual 
coala ara lowar 
Ikao llal rata. 

. _ ^ 

Uako «ala of Tka alala agaoay yaaa Ika Madleara •laodard Madlealtf |( a keapllal'a- Tka atala la ool 
Ikaraaaa aoal raporl ao4 ralroipaat Iva eoal raiinburaa- ooly ...la (o abova aooaldarloc a.y 
Coklrol ma*l ayala. elu. aa Idako-apael (la Ullallok ika cap. II allar.al II."| 

• ap 1. 11.11 tkir.aa... I.n.llo. I.dlaa. •.„ .S;.rb Ik. Ik I"! irM m., 

baaad o. a .alfklad .allo.al a«ara(a of koapl- addlllo.ol bo o««pal Ud i, * 

lal aoala aro iol U lo.r aoal ara gaa, ,oala. II aal- adopllk. Mad «,. 

««•. ll! a'Stu.'f.) TIlii'!*fl'" '*• "•••— "•<•! ••«• "P. Madlalld pro,,.™ lor 

ytor a aoala plua lallalloa aop). koapllal mial admlalilrallva 

roliabarao Ika aonalalaaay. 
alala acaaay lor 
Iko ovorpaymaal , 

Hlaaaaola Par Hoapllal-apaallla avarafa aAalaaloa rala la Madlaald Par'adiiil aalea Budial ravlaMal"" 

Aikalaaioa baaad oa 1(11 p.ymoal * admlaaloa d.la Iroad.d oaly ralo aoaalll.l.. Ika alala a.al 
iV,Z",t' ^*'» L*A boapil. -apa.llla alacla ika aallra ral»- a.l.l. I., . | 

ralo lor a.ory Madlaald .dn.l..|o.. bur.«..al lor a akarc-b.a.d pay.,. 

koapllal alay. la all aon-ladaral 
Tfcara aro ao koapllala, bul oam- 
rolroapaallva pllaaaa wilk bud- 
•^IxitMBla. gala la voluntary. 
• lata atay adopt Dao 
•yalam la Ika aaar 
lulura. 



EXHIBIT i.ii onMEKr ntosracTivB pavmekt SYsrnai ii itatbs amd i TERmitMir ooKriNUEo 
■•••............. 

"•*•"' '?'?''■ Tll'Ii;!* '"'"!'•?!'*•*'•" '"••"'••"''" M.<l|ei|d Prsvld.ri tbiorb Th« d.U iMl.ndt 
mittlon (HCVPRO) and Mtdletid •(•»«) tal ••vtn riltti only dirr«r*ne« b«l- lo rtriii* lh« 

lhr«* rsr |«n«r«l •dmltilom In thro hoaplltl ...n rat«( and iytlam b» addlnc 

paar (reupai ena aaeh for malarnlly, naonalal actual •ealt. additional rata 

inlanilva eara, and naoborn admliilonii and a ealaiorlaa 

par dlar* rata (or all admlaalont ovar It daya. 

NEVPRO prior approval all non-amarf ancy admii- 

alona and all ra-adml aaleni . Tha alata alto -> 

avaluataa aamplai of eaaai. All hoapilala mual 

comply wllh ayalam. 

Orafon p«r dia- Ho.pl I.I ..p., | f la rial a>ara|a ralaa par M.dl- Madlcald "Arihiindii Tha'itiii'piini'ir 

Human Raaoureaa. Ullllttllon ravla., which |al. Ju.lUy .yilam In July, 

rocu... on .dmLaloni. I. conduct. d. r.clll- dUr.r.ncaa ba- llil. Coal oui- 

llai ar. paid on Ih. b.ila of utilltallon t»,.„ actual ll.r. -Ill b. 

aummary . p.ymant. p., dlichar..! plua .n .0,1. .„d p.r pa.a.d throu.h. but 

'"' " '•««•'• All ho.pltal. ara Includ.d. dl.charj. p.y nol LOS oulll.;.. 

mtnl, .nd ara Tha ly.lam will ba 
than ralmburaad. tailad In fabruary- 
' March, Ids, In 

conjuncllon wllh 
Iha currant ay.lcm. 



states by unit of payment. It also includes a brief description 
of each system, which payers are covered, the extent of risk- 
sharing by the providers, and any future plans or relevant com- 
ments. (1) (In order to demonstrate the diversity and complexity 
of such systems, the next two sections of this chapter provide 
information on ten specific state Medicaid prospective payment 
systems . ) 

Finally, given the definition of PPS used in this study, 20 
states and territories were not using a prospective payment 
strategy for in-hospital cost containment in late 1984. All 
these states continued to use the standard Medicare retrospective 
"reasonable cost" payment system with the exception of Connecti- 
cut, which had a "prospective" system that involved no real risks 
to hospitals that incxirred costs above the prospective rates. 
These states 2md territories are as follows: . *. 

• Arkansas (desired a PPS, but no action yet) 

• Connecticut (had a prospective rate based on hospitals' 
historical costs, but hospitals were reimbxirsed at the 
end of the year for costs that exceeded the rates) 

• Delaware 

• Hawaii (A proposal to establish PPS for acute and long- 
term care facilities was rejected by the regional 
Medicaid office on grounds that cost reimbursement 
would be more efficient; rates were frozen for FY 85 
and the state planned to study PPS for future 
implementation. ) 



(1) By risk-sharing, we are referring to the part of the cost 
of delivering care that is not covered by the rates, budgets, or 
ceilings set by the state. If the cost of delivering care is 
above the per diem or the rate, some states make an adjustment in 
order to pay the hospital part of the cost of care, while others 
make no adjustments and hospitals are forced to eibsorb all of the 
difference. 



16 



• Indiana 

-•' • Louisiana (was considering PPS, but no action yet) 

• Montana 

• New Hampshire - ' -"'■ 

• New Mexico (was interested in DRG, but felt it was not 
easily adaptable to Medicaid) 

• North Dakota (was considering DRG) 

• South Carolina (had a PPS, but only for nursing home 
services) 

■- • South Dakota (was considering DRG) 

• Texas (did not think PPS would be worth the effort) 

• Vermont ■ ■" 

' m West Virginia (had a PPS, but it did not include Medi- 
caid; the state intended to include all payers by 1986) 

• Wisconsin (used a PPS, but the system did not involve 
Medicaid) 

• Wyoming (has inquired about a state rate commission) 

• District of Columbia 

''• • Puerto Rico (wanted a PPS, but no action yet) 

• Virgin Islands. 

2.5 Overview of Ten State Medicaid PPS Initiatives 

The ten state Medicaid PPS initiatives described in this 
section illustrate current trends toward PPS, as well as the wide 
range of alternatives being used by state programs. These state 
plans were selected to reflect the diversity of PPS initiatives, 
as well as some of the most recent developments. For example, 
four states — Ohio, Pennsylvania, Utah, and Washington -- are 
profiled because they have recently implemented or are about to 
implement DRG-based systems. At the same time that these 



17 



descriptions of DRG-based initiatives point to the influence of 
Medicare PPS on Medicaid, they also highlight the ways in which 
the Medicare model has been modified to meet particular state 
conditions. Similarly, we selected some states to illustrate 
other types of prospective payment systems — budget review and 
rate-setting, selective contracting, and revenue caps. 

The ten descriptions are based on interviews with HCFA 
federal, HCFA regional and state Medicaid administrators, as well 
as a review of available documents (e.g., state plan descriptions 
and regulations, recent contract studies, journal articles on 
PPS, the National Governors Association's tracking system, and 
reports from hospital associations ) . ( 1 ) The extent of detail and 
types of evidence collected varies across states, depending upon 
the type of initiative and availability of information. 

Exhibit 2.3 provides an overview and comparison of key 
elements of each of the ten state PPS initiatives. This exhibit 
contains the following information: (1) brief description of the 
PPS initiative; (2) year the current PPS initiative was 
implemented; (3) types of payers included in the initiative (e.g. 
Medicaid-only, all payers, etc.); (4) agency/commission charged 
with administering the initiative; (5) whether participation and 
compliance are mandatory for providers of services to Medicaid 
patients; (6) the prospectively set unit of payment; (7) the 
extent to which the initiative involves risk sharing on the part 



(1) Documents reviewed are listed at the end of this chap- 
ter. "^ 



18 



■XHIBIT t.l OKTAILCD DUCRimONS Or I* STATI fPS INITIATIVES 



llal* 



Dtierlpllon el ttt Ullltll** 



Admlnlilar- Unit of Eilanl of frovldor 
In* Afoiicv PoyiMnl Blak ShorliiR 



Ullllsolloa Coalroli 



Callfornlt aoUellvt Provldtr Conlroetin' 
-!>•> Profram (SPCP), undor (rhieh Ih* 
-Madiedd- alala nafollalaa aonlraela Indlv- 
enly Idually wllh allclbla provldara 
I* SI of III Haallh Paellillaa 
Plannlni Araaa (IIPPAil lo pro- 
»lda In-patlani hoapllal eara 
tor Madleald allflblaa. 



Cat Ifornia 
Madleal 
Aaal lanea 
Comnl ii Ion 
(CMACI 



■A 



All- 

Inclualva 
oar diam 






Contraela'i pricaa ara 
f I rni'-provldara abaorb 
Iha dlffaranea balwaan 
conlraetad priea and 
•elual eoali. 



Prior author laal Ion 
tor non-amarcaney 
aataai aanallona 
afalnal dumplnn. 






llllnola Proapaellva budaat and rala llllnoli lloiplial- 

'l*>l ravlaw lytlam which aalabllihai Haallh iDaelfle 

-Madleald mailniun affrafala budfala for PInanet par dlam 

o"ly hoapllala, aa wall ai hoiollal- Authority ralaa 
•pacific par dlam ralaa. 



Par dlcrn prieaa paid 
rafardlaaa of eoata 
of eara- -provldara 
abaorb the ditfaranee 
balwaan par diani rale 
and aelual ceala. 



£1 



Maryland 
-lift 
•All payara 



Proapaellva rale-aetlln* and 
budcal review ayalem which eala- 
bllahei relea per unit of aervlee 
bv eomparlnii direct eoata of eeeh 
teelllly with evarege eoata within 
peer irouplnca. Hoapllala have 
the option of requealinf full rale 
review le ealabllah a new achadule 
of ralaa or lo retein eurrent relea 
edjualad for Inflation. 



U 



lleellh 

Service 

Coat 

Rev lew 

Coomlaalon 



Plied per Hoapllal abaorba Iha 

unit of difference between 

aervlee aalabllahad rale end 

relet teluel eoata. 



In IITI, Merylend Im- 
plemented Outranleed 
In-pallant Revenue 
Syatera (OIR) et e 
aupplamenlary way 
of rewardlnf hoapl- 
lala that dacreaaa LOS 
and Intanallv of aar- 
vleaa within dlagno- 
ela eateeorlea and 
panaliilnf Ihoae that 
do not. 



Maaaaehuaetla 
-IIIO 

-All payara 



Proapaetlvely-determlned revenue Maaaaehu- 
eao eeleulalad for eeeh hoapllal aatla Rala 
with each payer reaponalbla for Selllnx 
e proportion of the revenue that Conmlaalon 
corraaoonda to Iha proportion of 
the hoipllel'i aervlee* uaad bv 
lla aubaerlbera. 



Overell Hoaoltala able lo keep 
cep on eoata below approved 
hoapllal revenue limit entitled 
revenue le retain the belance 
•a diaeret lonery In- 
eocnei If revenue cep 
la eiceeded they muat 
cover ahortfall. 



A revenue reduction 
factor la built Into 
the eap lo foaler 
produellvl tyi auball- 
lullon of eulpellent 
for Inpallent cere 
end reilonel planning 
of cere la eneoura<ed. 



DMIBIT I. a iMTTAiLao ocaoRimoNt or I* iTATi rn initiatives atKriMieni 



Allonbl* C:««l«/r*yiMiil 
A4|«iliMali 



■■•Imlaaa 



rr««l4ar Oranplaci ralvr* riant 



Olkar Motaa 



CA CompalllUa bld<lln( pra- 
aaii allkla aaak HPPA for 
aaalraalad aarvlaai. 
Annual rtvlaw of aan- 
Iraali Uriatloa a4- 
Juitmanl 



rravldara allkla 
II HPPAt (pra- 
domlaaaliy rural 
araat) aialudad 
fram (rCP laaa 
aalaali Iraa- 
alaadlai payakl- 
alrle kaapllali 
k akildraa'a koa- 
pilalt aaaludad. 



Paar (rouplaf aya- Padlalrle dlafaotlt asdlnf 
lam uaaa alallall- ayalam raaaally davaleptd 
aal alualar aaaly- allk III dltnastla aala- 
ala baaad aa alaa, (orlaa aalabllaklai oaa 
■ aatrapkla loaalloa. par 4la<a rata par diafaoalt 
laacklac aarvlaat, aala«ary. Tkia aill aaabia 
aoinplailty, ala. akildraa'a kaapllala lo 
uMfa la PPI. 



Conlraali undar SPCP ea»ar 
10% of alala McdI-Cal bud- 
(al lor la-pallaal aara. 
II of lit lirrAa eonaldarad 
loo aparaaly populalad lor 
IPCP-- Ikata araat ramtia 
undar faa-for-aar«lea allk 
ilrlacaal aoti lavlag la- 
aaallvaa. 



IL Ravaaua aap utat btta 

yaar aoilt Irandad for- 
■ard utiaf fcotpllal 
atarkal katkal lafladoa 
akaafaa la valuma aad 
(roMlk la madleal 
laabnoloiy. 



Raaaally paatad la(lalalloa 
Mould ttlabllak a aoalrael- 
laf ayilaia tiaillar lo 
Callforala'a irCP lallla- 
llva, affaellva la till. 
Hotpllalt Mllkia tpaalflad 
ragloaa naHld aan^all llvaly 
aacellala aoalraalt lo pro- 
*ldt tarvlaaa. 



■lata raaaally latlad a 
Primary Olafaatli Groupt 
iPOa) tyttam la tavaral 
kotpltalt la Ik* Chicago 
araa, timlltr le Mtdlctra 
WO lytlam but ollk laoar 
caltforlaa (III alatilfl- 
tallaaal. Pno tytlam It 
aipaclad lo ba abtorbad 
by propotad conlraallaf 
ayalam. 



lalllal ralaa baaad apoa dl- 
rael daparlmaalal aapaaaat 
ladlraal aipaatai alloea- 
lad by daparlmaal bad dabi, 
aktrlly, a payar dllfaraa- 
llal aad a eapllal faelll- 
llat allowaaea. lyalam In- 
•ludai aa laflalloaary 
ad|uilmtnl for koapllala 
aol raquaalinf full rala 
ravlaM. 



Paar groupla(a ara 
utad la Idaallfy 
faalllllat xkota 
•otit ara 'uaraa- 
tonabla* comparad 
lo timllar 
aarvlaat. 



At of I*/I/I4, Ika alala 
addad a DAO-battd alamaal I 
Ika rala ravlao tytlam, 
uadar wkleb Ika ilala alll 
pay lte« af avaraga tn 
Hadlaald LOS, vllkla fl»aa 
ma alaatllltalloaa. 



Hoipl lal •tpaci r le rava- 
aua cap ealeulalad utlagi 
pravlaui ytart approved 
ravaaua limit, laflalloa- 
ary aotI laeraaiat, DM 
caia-mli ad|utlm*ali lo 
voluroa, laekaoloiy ad- 
Juaimaal and cotit of 
aa* proframt. 



Sonw kotpltalt 
tbal did not pra- 
vloutly hava tlua 
Crott t|ratm«ntt 
(IIA II) not laclu- 
dad. 



Nona 



If tka Madleara valvar la 
aot ranaaad la lilt, Ika 
alala will aaad lo maka 
contldarabia ckangat In 
Ika oparalloo of tkIa all 
payar tyalam. 



Rapid coal Ineraatat (par- 
llaularly amployaa madleal 
Inturancal A kl(k eoncan- 
Irelloat of tacknolof let 1 1 y 
oriented tetcklng hoepllelt 
larialy raapoailble for In- 
illallon of revenue cep. 
lavolvemenl of bualnett, 
koipllel Indudry, A Blue 
Crott creeled itrong luppoi 
•upporl lor tkc Inltletlve. 



•11.»><. 



■XHIIIT (.1 DBTAILBD DESCmmONS Or It ITATB PP» IHITIATIVCS 



aid* 



DatarlplUn af PPS Ullltllv* 



A<k«lB|a|«r- Oall af lilaat ef Pra«ld«r 
Inc A(aaay Pajrmaal Rlik Ihirlnc 



Ullllaallan Caalrala 



MIehlian Preip*ell»« bud(«l ravlao lytlani SItIa Hospital 
•■•it uadar which paymania ara baiad on Madleald bud(al 
-Madleald- proapaellva aillmalaa ol aaah Afanay Until 
only hoapllal'a budfal. batad on 
aalual and hialorleal eoila. 



raelllllaa ara nol ra- 
Imburaad l( Iha* (o 
ovar budial llmlli 
raealva Ineanllva oay- 
manl If undar I loill . 



Utlllaallon control! 
Inaluda aurileal 
aaeand opinion pro- 
iram and ravlaw of 
ambulalary luralcal 
procadurai. 



Obia 0AO-ba*ad ayalani xhlch drawt Slata 

"'*** upon Mcdieara DOn modal, uilni Daparlmcnl 

-Midlcald- DRO claitirieat Ion tyilam, but of Public 

o"l« ollh Ohio-ipaclfle inalthtt and Walfara 
a ptar frouplnf lyilam baitd 
prlmtrllv on an araa wafa Indaii. 



Par Paymanl eonaldarad 
patlani eomplala and full-- 
dlicharia provider! abaorb 
ditlaranca balaan 
ailabllahad Dm rata 
and aalual coila. 



Raaiknlatlona . oulllari 
and Iranirara earaful- 
ly ravlawad to aiaura 
appropr lalanait and 
quality o( earai laml- 
annual ravlax ol aach 
laelllly. 



Oklahoma 
-III1 

-Madleald- 
onlv 



Pretpcellva rata-iallln| tystam Ralaa and All- 

Mhlch aitabllthai facility- Standardt Inelualva 

•pacific par diam ralmburaamant Advliory par diam 

rata lor In-pallanI hoipilal Comnlltaa rata 
• araleaa . 



Moapllal abaorba 
diffaranea balwaan 
aatabllthad rata 
and actual eoita. 



Maalmuni ralmburiabla 
IDS I* daya & no ra- 
a<knlialoni parmltlad 
within !• daya ol dla- 
eharfa (xllh caia by 
eaia aicaptlonil. 
Raalaw comnltlaa da- 
larmlnaa If oallanta 
raoalaa Inappropr la ta 
laval of eira. 



Pennsylvania 
-1*14 

-Madleald- 
oni V 



mo-basad proapaellva paymanl 
syttara Mhleh draws upon tha 
Madicara DRO modal, usinc 
DRO elsssi f Icat Ion systam 
but with Pannsyl vanla-spaci - 
fie walfhis and eomolaa paar 
(roupinf system. 



Stale 

Medicaid 

Araney 



Per 

pel lent 
dl seharia 



Peymcnl considered 
eomplele and full-- 
provlder absorbs 
diffaranea between 
the aatabllshed 
DRTI rata and actual 
coals. 



HosDilale estsblish 
utlllaallon review 
eonmllleea and plans. 
Review committee con- 
duets admissions and 
outlier reviews, end 
medical care evalua- 
tion studies to as- 
Bure Quality of care. 
Reedinleslon w| thin 1 
days does nol quality 
as separate edmlsslon. 



KmiiiT 1.1 DCTAiLED MacaimoN* or i* statb rrs iHiTiATivea cxiKriiiuEni 



AlUmbl* Ceili/P«)iMal 
A4|iialiMalt 



■■•lualoai 



rra«l4«r OrsuplBfi Pulara Pita* 



Olkar Nal«* 



HI CcplUI A l««ehl«( aotlt 
laduiitd «t *pparllaa«4 
Mi*4 (meual. Hl|h M«4I- 
•tl4 *alum« •dlutlnwalf. 



■•kibllllallv* and 
payehUIrl* ladl- 
lalloa* •■•ludad 
(ram Ik* •ytlMi. 



Hoa* 



Plan lo laiplamaal DNO-baaad 
ayalam la l/ll. dhlab alll 
ba almlUr la Madlaara. 



ai Capital aaala laaladad 
la owl paynMal-- par- 
aaalafa alleHaaaa baaad 
aa ralie af aapllal ra- 
lalad aaili la aparallnf 
aeali. Paal 1 1 ly-apaal* 
(la atfjualmaal far haa- 
pilala aarvlng •dlapra- 
parllaaala auinliari*. 
T.IK laMallan adjual- 
maal aap. 



Praa-alaadlaf ra- 
kabllllallva (aa- 
llllllaa. Ian( 
larm aara batpl- 
laU aad SHP/ICP 
ualU aieludad. 



Paar irauplaf baaad 
o* ■•(« ladai af 
SMSA Bhara kaapllal 
laaalad. taparala 
paar fraupa aala- 
blUkad far laaak- 
l«t A ahlldraa'a 
koipllala, rural 
koapllala a»ar !•• 
I>ada and raral kai- 
•llala uadar !•• 
badi. 



DIUl-baaad ayalam le ba 
phaaad la a«ar t yaara by 
blandlng keapllal-apaalf la 
ralaa adlk paar iraup prl- 
aaa la daallalac propar- 
llana (tt/lt (er Pv II. 
II/TI (ar PT II). By T/ll 
kaapllala la ba paid par 
diaakarfa baaad on avarafa 
aoat af paar (roup. 



Ona premlaaal coaaldaral Ion 
oaa lo maka PPS almllar 10 
Hadlaara ayalam ae koapl- 
lala aould ka«a almllar op- 
arallant acrott procrania. 
DAO ayalam xaa davalopad 
In-houaa wl Ik aaalilanca 
from a aoniulllng firm. 



OK Ralaa aalculalad uainf 
Uadlaara aoal raporia 
tar aaak kotpllal ad- 
lualad (or laflallaa. 
Ha adjualmaol (or Ika 
aurraal aalandar yaari 
aubaa^uaat yaart aa 
aallmalad laflallaa 
adjualmaal aaad baaad 
oa liCPA'a Biarkal baakal 
ladai. 



Nona 



Hoaa 



Prallmlaary raporia Indl- 
oala Ibal alaaa Implamanla- 
lloa o( praapaallva ayalam, 
ao laaraaaaa la la-pallani 
Madlcald aoala kava occurad 
(aaoiparad la a 40% Inaraaaa 
la Madlaald coala durlag 
Ika pravloua 1 yaara). 



PA lUllar lo Madlaara DM] 
allawabla aoala. Capllal 
bulldlaf A (lalura eoala 
paiaad Ikraugk. Ilala- 
wlda ralallva Malfkia 
updalad aaaually uainf 
aoal raporia, lavolaaa 
aad audita. Ilala uaaa 
aoal raporia lo aaleu- 
lala rata o( laeraaaa 
(or paar freupa. 



Payaklatrla, dru( 
aad alaokal rakab- 
llltallaa ualla o( 
(aaaral keapltalad 
aarallad la Madlaal 
Aaalalaaea Profram 
ara aialudad. 



Complai paar (reup- 
ln( ayalam baaad en 
(our aoneaplai 
laaaklng alalua, 
Madlaal Aialalana* 
voluma, aadronmaa- 
lal akaraalarlallea 
and kaapllal aaala. 
VInal ranklaf a( 
haipliala raaulla 
la I paar Rraup|a«a. 



Day oulllar aialkodalagy lo 
ba davalopad (or aaak ORO 
(aaaapl Ikoaa ralalad lo 
burn and naonalal aara) 
alan( llnaa aalabllabad 
(or Madlaara program. 



Hoapllal paymania mI 1 1 ba 
II/TI bland of paar group 
ORO and koipl lal -apiel ( le 
ralaai Ihli bland mI 1 1 ba 
11/11 Ika aaaond yaar and 
lim ORO baaad Ika Ikird 
yaar. 



KxiiiaiT t.i orrAiLD) neacRimnNS or i* state pps initiatives 

SI*U D«t«rlplloii or rrs Ulllillt* A<kiiliilal«r- Unit of Eitanl of rravldar UlltUallen C«iilr«li 

luf Afaitcf P«fnMiil RItk Sharln* 

Ul*h DRO-bdcd proiptellv* ptymtnl lyi- Slat* Par II aelual eoala laaa Prior author Italian 

-Itil lam drawa upon Madlcara DRO modal Madleald pallani than DRO larfol rala, and coneurrani ravle* 
-Madlcald uainf Iha aamo elaaal (leal Ion afi- Afanev dlaeharfa alaU pay) <oala plut ohleh arara pari ot 

lam (or diagnoala and ralatlva ie« (not lo aioaad lar- pravloua ayalam, ara 

Malfhli, bul utlnf Madleald elalma gal rala)| l( eoala ai- bulll Into aurranl 

data lo raduea DRO paymant ratal. etad largat rala, atala ayilam. 

Actual paymanta lo provldara ara paya lar«al rala until 

caleulalad with a 'corridor' around coata ara tO% abo«a 

largal DRH rala. lariat rala, than ilala 

paya 10% o( additional 
eoala. 

»« '. f • ' . 



X *■ 



• >.»»■■ 



Waihlngton DRO-baiad proapaetUa paymant aya- State Par Payment aoneldered coin- H/A 

-l*l< tern, uilng Medleere DRO eleaald- Madleald patient plele end (ull-- pre- 

•Medlcald- cation lyilem, Melfhta end hoapltel- acancy dlaeharge vider abaorba dllfarenee 
only apeeldc evereie coat per eeie, between eatebllehed ORG 

with Medleere retea reduced eeroaa rale end actual eetla. 

the boerd to (It within atete bud- 
fat. 



■XHIBIT I.I Drr*ILID DeSCXIPTIONS OP 11 ITAT* PPt IHITIATIVU OnKriNUEDi 

Alloaabla Codi/rapMBl tialailcaa ' Pravldar Orauplaca fulMra Plaaa Olh<r Nalat 

A4|aalawa<a 

UT Utat Madlaar* DUO Cklldraa'a hoa- || i, anilalpalad Itili'l 

allaHtbIa aotla <ln- piiala aialudad alala-ipaal (la DRO 

aludaa paaa-lhroufh taatai Ikara ara pf Icing ij t lam ol 1 1 ba 

ol aapllal a*d adu- as fraa-(ta«4lag davalopad baaad on pro- 

aatloK aoala) payahlalria ar «|dar aod Inrorm.l Ion. 

rahabllllalloa (lata li alio eonildarlnf 

keapltala). moving la a •alacllva 

aenlraaling program. 



WA Oparallng, aapllal, and 
laachlng coal* Ineludad 
la tha txta rata. Coali 
paatad tkrougk Inaludai 
lalaraal, dapraalalloa, 
dadualloaa la ravaaua 
baaad on voluma af 
Madlcara/Hadlaald pa- 
llania, diraal madleal 
aduaallaa aoala. 



Payahlalria, ra- 
kabllllalloa, and 
ekildran'a koi- 
pllala aialudad. 



Tfcara ara plana lo 
aal up paar groupa 
laa yal, undarUad). 



II la aipaalad thai Ibli 
• yaianialll avanlually ba- 
aoma an all payar ayalam 
wllfc paar grouplnga. 



Unlll Implamanlal Ion of DRQ- 
baiad PPS, a rata aalllng 
and budgal ravltn lyalain 
•III oparala, ohlch In- 
volvaa ravlaw of major coti 
aantara wllhin aaek koipl- 
lal. Budgal leraani ara 
u(<d baaad on all hoipllili 
vllhln alala--eotli abovt 
lh« •eraani ara dl •• I lonad. 



of the health care provider; (8) the extent of controls over 
utilization (e.g., length of stay (LOS), review of service 
levels, readmission, etc.); (9) types of allowable costs within 
the PPS formula (particularly relating to capital costs); (10) 
types of providers or departmental units excluded from the 
methodology; (11) retroactive and/or annual adjustments to 
provider payments (especially for inflation, volume adjustments, 
and case-mix); (12) types of groupings of hospitals/providers 
(particularly in terms of peer groupings); (13) future plans and 
proposals for development of the state's initiative; and (14) 
other relevant notes /comments on the PPS initiative. 

A comparison across each of the key elements of the state 
PPS initiatives underscores the hybrid nature of Medicaid 
prospective payment systems. This extensive variation is 
explained in part by the flexibility extended to states in 
administering Medicaid, as well as by the unique conditions and 
problems found within each state. For example, as shown in the 
individual state descriptions below ( Section 2.6), the underlying 
influences and interests involved in the development of the 
Massachusetts prospective revenue cap system and the California 
contracting system were very different. in Massachusetts, the 
business community, hospital industry, and major payers (e.g. 
Blue Cross /Blue Shield) were substantially involved in setting up 
the prospective revenue cap and strongly endorsed this system. 
In California, the hospital industry and business community did 
not substantially contribute to the development of the selective 
contracting system, which is built around a competitive bidding 
and negotiation process that pits providers against each other, 

19 



and consequently actively opposed the initiative. As a result, 
the mechanics of these two systems provide strong contrasts. 

Four models emerge from the "basic descriptions" of the ten 
state PPS initiatives profiled in Exhibit 2.3: (1) DRG-based PPS 
initiatives (e.g., Ohio, Pennsylvania, Washington, and Utah); (2) 
prospective rate-setting and/or budget review (e.g., Oklahoma, 
Maryland, Michigan and Illinois) ; (3) selective provider con- 
tracting (e.g., California); and (4) prospectively determined 
revenue cap (e.g., Massachusetts). 

Since the development of the DRG classification system and 
the implementation of this form of PPS in New Jersey and in the 
federal Medicare program, the DRG-based model is being 
increasingly adapted to state Medicaid programs. The four DRG- 
based systems that are highlighted in this section illustrate not 
only the willingness of states to draw upon the DRG methodology, 
but also the perceived need to adjust this methodology to the 
conditions within each state. The models in each of these states 
differ from the federal Medicare model and from each other. For 
example, Utah, like Ohio and Pennsylvania, uses the Yale DRG 
classification system. However, unlike Pennsylvania and Ohio, 
Utah first uses the Medicare DRG relative weights to establish 
pricing and then employs Medicaid claims data to reduce actual 
payment rates. Ohio and Pennsylvania use state-specific weights 
to determine payment rates. These three systems also differ 
substantially in such areas as provider risk sharing, utilization 
controls, types of facilities excluded and peer groupings. 

The DRG-based PPS model, while increasing in importance, is 



20 



not the only or even the predominant prospective payment model 
now in existence in state programs. Many states, such as 
Maryland, Illinois, Oklahoma, and Michigan, have developed 
complex rate- setting and budget review systems which 
prospectively establish rates per unit of service, per day, or 
per discharge. A good example of this approach can be found in 
the thirteen-year-old prospective rate-setting and budget review 
system in Maryland. The Maryland system uses a quasi -public 
utility apjproach to hospital rate-setting, under which rates are 
prospectively set and then adjusted for inflation, volume changes 
and pass-through costs. An independent commission appointed by 
the governor, the Health Services Cost Review Commission, is 
responsible for overseeing development of the hospital-specific 
prospective rates. 

A third approach to prospective payment involves selective 
contracting. The most important example of this reimbursement 
form is the Medi-Cal program in California. Under this system, a 
state commission contracts with hospitals, county health systems, 
HMOs, and private insurers for the provision of pre-paid, in- 
patient medical services for the Medi-Cal eligibles. The prices 
are arrived at by negotiations and competitive bidding. A 
primary objective of this form of reimbursement is to introduce a 
level of competition between providers with respect to price of 
in-hospital services. 

In strong contrast to the selective contracting of the 
California system is a fourth major PPS model: the prospective 
revenue cap system found in Massachusetts. Under this type of 
prospective system, a revenue cap is negotiated for each 

21 



hospital. Each payer is then responsible for the proportion of 
the hospital revenues that correspond to the proportion of 
hospital services used by the payer's subscribers. 

In Exhibit 2.3, the "year of implementation" refers to the 
date in which the current PPS initiative went into effect. 
Across the profiled states, there has been considerable activity 
with respect to reimbursement systems, with all the states having 
introduced structural changes in reimbursement since 1980. 
However, it is important to note that many of the states profiled 
in this section have been using some form of PPS since the early 
1970s. Perhaps the most notable of the recent changes in the ten 
states illustrated here is the development of DRG-based systems. 
Ohio, Utah, and Pennsylvania have all adopted DRG-based systems 
since 1983 — and Michigan and Washington were planning to intro- 
duce DRG-based reimbursement in late 1984 and early 1985. 

Important feat:ires of any PPS initiative are the type of 
agency administering the initiative and the scope of authority of 
the agency. The two predominant "administering agencies" across 
the ten state Medicaid programs are independent rate-setting/re- 
view commissions and the state Medicaid agencies. All of the 
systems profiled had mandatory "provider participation and com- 
pliance;" that is, if providers were to serve Medicaid clients, 
they had to participate in the process used to establish prospec- 
tive payment rates and had to comply with the terms of payment 
for clients served. In terms of scope of authority, or "payers 
included," most of the ten state systems are Medicaid only. Two 
states, Massachusetts and Maryland, are all- payer systems. Sys- 



22 



tems that cover a larger proportion of payers are generally 
considered more likely to have a stronger impact on containing 
hospital costs. 

The ten PPS initiatives reflect considerable variation in 
"unit of payment" available to states. Four different units of 
payment are used (some in combination with one another): (1) per 
diem, (2) per service, (3) per patient discharge, and (4) overall 
revenue cap or budget limitation on the provider. 

The extent of "provider risk sharing" is an essential 
characteristic of prospective systems. Under PPS, providers 
generally ass;ime all or some part of the risk of providing 
medical services. All of the systems profiled had some form of 
risk sharing. In most state programs, the health care provider 
absorbs the difference between the prospectively established 
rates and the actual cost of providing services. In other 
programs, the providers are at risk for some portion of the costs 
in excess of the established rate and retain some portion of the 
payment if actual costs are below the established rate. For 
example, in Utah, if actual costs are less than the DRG target 
rate, the state pays costs plus ten percent (not to exceed the 
target rate ) ; if costs are in excess of the target rate , the 
state pays only the target rate until costs are 60 percent above 
the target rate, at which time the states begin to pay a portion 
(60 percent) of the excess cost. 

Consistent with the objective of containing the cost of in- 
patient services and with controlling the level and quality of 
medical services, many of the ten initiatives profiled have 
introduced "utilization controls." These controls have taken on 

23 



several different forms, including: (1) review of limits on 
length of stay (LOS), (2) restriction of readraission and transfer 
of patients, (3) review of patient service utilization, (4) 
review of quality of care, (5) estajolishment of state and 
hospital- level review committees, (6) prior authorization for 
non-emergency care, (7) incentives for substituting out-patient 
care, (8) incentives for regional planning and coordination of 
specialized services, and (9) sanctions against "dumping". 

The "allowable costs" and "exclusions" are important factors 
in determining overall stringency of PPS systems. In general, 
the fewer unallowable costs and exclusions, the more stringent 
the system in terms of controlling costs. In some cases, 
allowable costs are defined in terms of Medicare "reasonable" 
cost principles. Other states have defined their own criteria. 
Of particular interest in this area is the treatment of capital 
costs, depreciation, and teaching costs. Similar to "allowable 
costs," each initiative — depending, in part, on the type of 
system and the structure of the hospital industry — excludes 
different types of institutions. Typically excluded providers 
are free-standing rehabilitation facilities, psychiatric 
institutions, children's hospitals, long term care hospitals, 
SNF/ICF, and rehabilitation, and psychiatric units. 

Many of the prospective systems have some type of "payment 
adjustments" to improve equity of the systems across hospitals 
and to adjust rates on an annual basis. The ten state 
descriptions offer illustrations of some of the typical payment 
adjustments: (1) facility-by-facility adjustments for 



24 



"disproportionate" numbers of Medicaid patients, (2) annual 
inflationary adjustments based on the HCFA Market Basket Index, 
(3) retroactive adjustment of payment levels if the projected 
inflation rate is different than actual inflation, (4) peer group 
inflation adjustments, (5) technology adjustments, and (6) case- 
mix adjustments. 

"Provider groupings" is another distinguishing 
characteristic of prospective systems. The grouping of providers 
-- often into "peer groups" — is used to compare and develop 
price levels for services among similar types of institutions and 
to monitor levels of service utilization. These groupings are 
widely divergent across states; partly due to the type of PPS and 
the types of institutions included in the prospective system. 
Some of the defining characteristics used in determining peer 
groupings include: (1) teaching status, (2) geographic location 
(e.g., urban, suburban, rural), (3) number of beds (e.g., 100 
beds or more), (4) speciality hospitals (e.g., rehabilitation, 
children's, psychiatric), (5) medical assistance volume, and (6) 
level of hospital/ service costs. 

Finally, it is apparent from the ten profiled state plans 
that there will continue to be sxibstantial activity with respect 
to their reimbursement systems. In some cases, "future plans" 
involve refinements and adjustments to the current system; in 
others they involve the implementation of entirely new systems of 
reimbursement for providers. •»,:..■ 

The remainder of this section provides detailed descriptions 
of the ten state PPS initiatives. The descriptions provide back- 
ground on the origins and development of each system, a discus- 

25 



sion of key elements, examination of design and/or implementation 
problems, and information about plans for future development of 
the systems. 



:n , 



26 



2.6 State Descriptions 

CALIFORNIA 

The Selective Provider Contracting Program (SPCP) in 
California has been the focus of intense national interest as 
one of the major models for Medicaid cost saving and prospective 
payment; it has also been the center of intense local controversy 
ever since the program was implemented in 1982 to help avoid a 
major budget crisis in the state. 

Under this system, payment for in-patient care under Medi- 
cal is covered by an all-inclusive per diem, negotiated 
individually with eligible providers on a regional basis. 
Originally, these rates were negotiated by the Governor's Office 
of Special Health Care Negotiation (GOSHN), popularly known as 
the Medi-Cal "Czar". Since July of 1983, however, this function 
has been taken over by the California Medical Assistance 
Commission (CMAC). 

Under SPCP, the state is divided into a nximber of Health 
Facilities Planning Areas (HFPAs). Within each HFPA, the state 
has estimated the number of bed/days needed to provide Medi-Cal 
services for the coming year. Eligible hospitals within the 
HFPA were then allowed to negotiate with the state on a price at 
which they would agree to provide a portion of the projected 
services. The contracted price is firm; the hospital can 
certainly keep any difference if it manages to provide services 
below the contracted price, but it also absorbs any costs that 
exceed the contracted price. Contracts were negotiated with 
selected providers within each HFPA until the projected Medi-Cal 



27 



need was covered. 

Since the number of bed/days bid was higher than the 
projected need, the state was able to negotiate some very large 
discounts into the proposed prices. In many cases, this discount 
was as much as 30%. 

When contracts had been signed to cover all the projected 
need, the HFPA was considered "closed" — that is, with the 
implementation of SPCP, Medi-Cal patients must go to a contract 
hospital. There are sanctions against dianping; a facility must 
treat a Medi-Cal patient if it is under contract to do so. Non- 
contract facilities may now only provide emergency services or 
specialized services not available at contract hospitals. These 
contracts cover approximately 90% of the state Medi-Cal budget 
for in-patient hospital care. Of the 137 HFPAs in the state, 56 
are "closed" under the contracting system. 

The remaining 81 HFPAs are considered to be too sparsely 
populated to assure access if contracting were implemented. 
These areas remain "open" to fee-for-service, with cost 
incentives calculated through a peer grouping system, adapted 
from the categories used for other purposes by the California 
Health Facilities Commission. This peer grouping system uses 
statistical cluster analysis based on seven or eight major 
variables, such as complexity, size, geographical location, and 
teaching services, to classify hospitals. 

Under this system, called the State Plan Method, cost saving 
incentives are incorporated by using the allowable cost per 
discharge for the peer group as the unit of comparison. All 
facilities in the 60th percentile and below are considered to be 

28 



operating efficiently and can obtain adjustments to their per 
diem. All facilities operating above the 60th percentile are 
assumed to be operating inefficiently and are penalized in pro- 
portion to how far they are operating above the 60th percentile. 

Psychiatric units of general acute care hospitals under 
contract are covered under SPCP, but the free-standing 
psychiatric hospitals are exempt from the contracting system. 
Alcohol rehabilitation services are not covered by Medi-Cal and 
therefore do not come under either SPCP or the State Plan. 

Children's hospitals are also exempt from the contracting 
system. Originally this was done by executive fiat, later 
formalized by a two year legislative exemption. This exemption 
was based on the expectation that the children's hospitals would 
use the time to develop their own system of prospective payment. 
Accordingly, the children's hospitals have joined together in a 
contract with a major consulting firm to develop a pediatric 
diagnosis coding system. This projected coding system will 
create 215 diagnostic categories that are analogous to, but not 
the same as, similar diagnostic categories under the Medicare DRG 
classification system. There will be one per diem rate for all 
hospitals within each diagnostic category, but adjustments for 
case-mix will be made for each facility. These case-mix indexes 
will be adjusted every six months, to avoid diagnostic category 
"creep", cost shifting, and other gaming strategies. 

Utilization controls are not a specific part of the SPCP 
program, but a requirement for prior authorization already 
existed in the state and has been retained and expanded in light 



29 



of the new approaches to cost containment. More stringent limits 
in other areas such as eligiblity, ancillary services, drugs 
listed in the formulary (i.e., acceptable under Medi-Cal), and 
other areas have also contributed to control of volume, but are 
not part of the contracting program per se. 

Most contracts with providers under SPCP do not have a 
formal termination date, but they do call for review after one 
year. As of approximately a year ago, the first of these reviews 
were in process, and many of these same providers should be 
coming up shortly for a second review. 

A Program Evaluation Team (PET) is now being put together by 
the state to evaluate the effects of the entire SPCP program, but 
it is too early to determine any useful information about this 
effort. 

ulinois 

Illinois pays for Medicaid in-patient hospital care through 
a fixed per diem price system. This current system, implemented 
in October 1981, is based on a fixed dollar appropriation at the 
state level. The state projects the prospective bed/day need for 
its coming fiscal year, and examines hospital costs for the 
individual facility's previous fiscal year (these two types of 
fiscal years may not be the same ) . On the basis of these 
factors, a hospital-specific per diem is calculated that is 
intended to cover the projected need without exceeding the state 
budget. By July 1, 1982, this system had been phased into all 
hospitals in the state. 

Under this system, the risk is on the hospital to provide 



30 



services at the per diem rate. But since the rates are based on 
a budget review, if the hospital spends less, its rates go down. 
If the hospital exceeds its portion of the state budget, the 
overrun is deferred for repayment to the state in an unspecified 
future year. According to the National Governor's Association, 
these types of features have contributed to the failure of 
previous PPS systems in Illinois. 

Recently, as a result of state legislation calling for the 
Medicaid agency to "try" to decrease utilization, the state 
embarked on a pilot project in the Chicago area to experiment 
with a DRG-like system. The system is based on an Illinois- 
specific classification system of Primary Diagnosis Groups (PDG) 
which consists of 218 classifications, fewer than the Medicare 
system. Payment is based on the discharge that is most resource 
intensive, not the admitting diagnosis. According to regional 
HCFA officials, other than those two factors, it is very similar 
to Medicare DRG. t- . -- - ■ 

Under the PDG system, hospitals contract with the state to 
provide services at 95% of the normal Medicaid rate. The 
hospital is then forced to reduce utilization as the only way to 
continue to cover operations. The participating hospitals were 
competitively chosen in 1982. Sixty hospitals sxibraitted bids, 
but only seven were chosen. These hospitals continue to submit 
claims under cost-based reimbursement, but are then paid only the 
reduced rate. If they can provide services under the contracted 
cost, they can keep the difference. .- ^ 

According to regional HCFA officials, the ink was barely dry 
on these contracts when the legislature passed new legslation to 

31 



clarify and expand on their previous statements. This legisla- 
tion mandated the state Medicaid agency to enter into contracts 
with hospitals. The state agency iinmediately began examining the 
waiver proposal that the State of California had submitted in 
order to implement its Selective Provider Contracting Program 
(SPCP), and soon submitted a very similar proposal to the re- 
gional office. 

This system as proposed in Illinois would establish geo- 
graphical regions within the state, and would project an esti- 
mated number of Medicaid beds needed for that area over the next 
year. All hospitals in the area would then receive an offer to 
bid "on the beds. Once contracts were awarded for 100% of the 
estimated bed need, hospitals without contracts would be unable 
to receive payment under Medicaid for anything but emergency 
services . 

The HCFA Regional Office was reviewing this waiver proposal 
in late 1984 and, although they had some reservations about it, 
they expect to approve it eventually. The reservations included 
the following: 

o the proposed payment mechanism was too unclear to be 
implemented, and the budget constraints the system 
would operate under were not spelled out; 

o the proposal requested the waiver of many utilization 
review requirements, some of which are in waivable 
sections of the Social Security Act and some of which 
are not; it is unclear whether the latter can be 
legally waived; 

o there was no arrangement for payments to out-of-state 
hospitals which serve Illinois Medicaid recipients. 

The proposal also requested the waiver of freedom of choice 

for Medicaid recipients, and included a requirement for prior 

32 



•f 



authorization. The proposed system was slated to go into effect 
in March of 1985. 

Since the passage of the bill authorizing this SPCP 
program, there has been little or no interest in the PDG program. 
The PDG program will probably be absorbed into the SPCP program, 
with little or no further development in the direction of a DRG- 
type system. Illinois seems to have wholeheartedly adopted the 
California model of selective contracting. 

MARYLAND 

The State of Maryland uses a rate- setting system with a 
quasi-public utility approach to regulation. Implemented in 1971, 
this was the first rate-setting system in the nation. The system 
includes all payers in all non-federal hospitals, and includes 
the costs of all services associated with hospitals. For 
example, there are six hospital-connected nursing homes in the 
state. The rates for these nursing homes are set under this 
system, but no other nursing homes fall under this jurisdiction. 
Medicare and Medicaid participate on an experimental basis. By 
law, this payment system may not jeopardize the solvency of any 
hospital in the state. 

The rates are developed by the staff of the Health Services 
Cost Review Commission, approved and supervised by the Commission 
itself, which is an independent body consisting of seven members 
appointed by the Governor. A majority of the members of the 
Commission may not have any connection with the management or 
policy development of any hospital or related institution. 

The prices of services for hospitals were originally set 

33 



through a system of rate review, which is discussed below. In 
subsequent years, these original rates are simply adjusted for 
inflation, for most facilities. To place further limits on the 
providers, the state added a program of Guaranteed Inpatient 
Revenue (GIR), which offers incentives to decrease LOS. At the 
present time, the vast majority of discharges in the state fall 
under GIR. The GIR system is also covered in more detail in 
subsequent sections. 

On October 15, 1984, the state added a utilization 
limitation based on Medicare DRG. Formerly, the state reimbursed 
hospitals for treatment of Medicaid patients up to a limit of 
twenty days across the board. Under this new provision, the 
state will pay 120% of the average Maryland Medicaid LOS for that 
DRG, based on state-specific information developed by the Cost 
Review Commission. This novel use of DRG classifications for 
utilization control only will not affect rate setting ger se. 
However, it is clear that many cases will be limited to 
substantially below twenty days of reimbursement, which puts new, 
more stringent limitations on the providers. 
Rate Review 

The implementation of rate setting originally required a 
detailed review of the budget of each hospital in the state, 
which is described in this section. For most hospitals, the 
original rates are only updated, as described in the following ■ 
section. However, under the law, although the option is seldom 
exercised, every provider has the right to have its rates 
recalculated under the original methodology. 



34 



The initial set of rates per iinit of service was 
developed through the use of the Interhospital Cost Comparison 
(ICC) methodology. Facilities were required to submit data for 
the various revenue producing departments for base and budgeted 
years, in accordance with a uniform accounting system. Approved 
revenues are based on direct departmental expenses, indirect 
expenses allocated by department, bad debt, charity, a payer 
differential, and a capital facilities allowance. In addition, 
peer grouping is used to identify facilities whose costs are 
unreasonably out of line with similar facilities in the state. 

The peer grouping compares the actual total direct costs of 
a facility with the average of the peer group. The Commission 
then sets rates through a very complex formula that includes not 
only the information from the budget review described above, but 
also variables such as location, case-mix intensity, scope of 
services, extreme variations of size or payer, and individually 
negotiated agreements with many facilities. The complexity of 
the process results in formulas and rates that are de facto 
hospital-specific, although this is not an explicit aim of the 
system. - 

The Capital Facilities Allowance (CFA) includes two 
categories: buildings and moveable equipment. The building 
allowance is the larger of 20% of the replacement cost or the 
current cash requirements of a building, including debt and major 
maintenance. Moveable equipment is compensated based on a survey 
of replacement value that is used to establish a statewide 
moveable equipment allowsuice per bed. In addition to the CFA, a 
working capital allowance of 2% is factored in as an allowable 

35 



cost. 

The use of replacement value, as opposed to historical 
depreciation is an innovative facet of the Maryland system. This 
method has a number of advantages. It provides incentive to 
closely monitor costs and discourages replacement where there is 
no need. A fixed/ variable cost ratio, generally 40/60 for 
routine cost centers and 60/40 for ancillary services, guarantees 
fairness during significant volume changes. The replacement 
allowances are comparatively generous, however, and there is no 
link to health planning — CON approved projects are merely 
passed through. 
Inflation Adjustment 

Each hospital always has the option of requesting a full- 
rate review to establish a new schedule of rates, but this is 
seldom exercised. Usually the hospital chooses to retain the 
current schedule, adjusted only for inflation. Adjustments are 
made for salaries and fringe benefits, and food, supplies, 
utilities, and other operating expenses. There are three parts 
to the adjustment formula: 

(a) a retroactive adjustment, if actual inflation was 
different from the projected rate; 

(b) a price- leveling adjustment which brings the rates to 
where they would have been if inflation had been 
projected correctly; and 

(c) a new inflation projection for the coming year. 
Costs can be passed through if they are mandated by state 

or federal government and affect all hospitals (such as changes 
in the minimum wage) or if there is an increase in costs greater 
than the CPI which affects the hospital industry more than other 

36 



industries (such as increases in FICA) . 

Volume for the prospective year is established at a level 
equal to the actual volume for the current year. Different 
fixed-variable cost proportions have been established for the 
routine and ancillary areas as well as for different magnitudes 
of volume changes. „^ 

Guaranteed In-Patient Revenue System 

Systems based on rates per service are often soft on 
utililization controls. In an effort to control these factors,, 
around 1978 the state implemented the Guaranteed In-patient 
Revenue (GIR) system as a supplementary method of rewarding 
hospitals that decreased LOS and intensity of services within a 
diagnostic category — and to penalize them if they did not. 
This system is unusual in its attention to intensity of services. 
Under GIR, the average hospital charge per diagnosis and 
per payer is determined per discharge. This average charge is 
adjusted for inflation, plus a 1% factor for growth and 
technology. The hospital is assured, at the minimum, of the 
level of revenue this formula projects. At the end of the year, 
revenue to the hospital under GIR is compared to what revenue 
would have been under the regular state-set rates. If the 
hospital charges are under the GIR projection, then the hospital 
receives the variable cost portion of the savings as a reward. 
But, if the hospital exceeds the projected GIR, the state will 
recoup the additional funds from the hospital in the following 
year. 



37 



c 



MASSACHUSETTS 

Some analysts consider the Massachusetts pps, popularly 
referred to as Chapter 372, to be the antithesis of the program 
in California. These two programs demonstrate the diversity of 
major policy options open to states in their efforts to contain 
the spiraling costs of medical care. David Kinzer, in a recent 
article in the New England Journal of Medicine, noted that: 

a large question for the nation to consider is whether 
It wants a health care system, with the machinery for 
controlling costs, that is modeled on the system in 
California or the system in Massachusetts . . . ( I ) t is 
hard to imagine how the divergent approaches taken by 
California and Massachusetts will ever be reconciled. 
The final decision could go either way, but it seems 
unlikely that the result will be something in between. 

Like most states, Massachusetts previously had a cost 
reimbursement system. Rapid cost increases hit particularly hard 
in the Massachusetts area, which has a high concentration of 
technologically-oriented teaching hospitals. The recent 
recession, which resulted in many uninsured patients, has been 
worse in Massachusetts than in many other states. In addition, 
the state Medicaid budget has not kept up with the cost of 
hospital care because of mandated restrictions in the property 
tax rate similar to Proposition 13 in California. Finally, the 
Omnibus Budget Reconciliation Act required Medicare payments to 
be below the level of Medicare costs. Given such conditions, 
hospitals will often pass the cost increases on to paying 
patients — especially those covered by third parties. 

Prior to the implementation of Chapter 372, in an effort to 
contain costs that were rising at a rate of 15 to 20 percent a 
year, the Massachusetts Hospital Association and Massachusetts 



38 



Blue Cross signed an agreement establishing a revenue cap — the 
maximum allowable cost, or MAC — on annual Blue Cross payments 
to each Massachusetts hospital. Under Chapter 372, which was 
initiated in 1980, the MAC concept was extended to cover all 
payers . 

Under this system, a cap is placed on the revenue of each 
hospital. This cap is calculated individually for each 
institution. Each payer is responsible for a proportion of the 
revenue that corresponds to the proportion of the hospital ' s 
services used by its subscribers. A revenue reduction factor is 
built into the cap to foster gradual improvement in overall 
productivity. Volume reductions are encouraged, as is 
substitution of outpatient care for inpatient care. Regional 
planning is also encouraged by allowing hospitals to distribute 
among themselves the revenue allowance for services that are 
terminated as a result of mergers or cooperative arrangements. 
Hospitals that are able to keep their costs below the approved 
revenue limit in a given year are entitled to retain the balance 
as discretionary income, but if they exceed the revenue limit, 
they must cover the shortfall themselves. 

■ : To date, this reimbursement system is perhaps the most 
stringent revenue- limited, prospectively budgeted hospital 
financing system instituted by a state. However, it is still 
less stringent than Medicare DRG. This is an important 
consideration to supporters of Chapter 372, and it has made 
obtaining and keeping a federal waiver crucial to the success of 
the negotiation process. 

The Massachusetts approach can be described as a kind of 

39 



economic franchise that, in effect, can reinforce and sustain 
existing cost and price disparities between hospitals. The state 
tells each hospital a year ahead how much revenue (therefore, 
expense) it can have within tightly constrained limits that are 
set at "inflation-minus" increments building from base year 
information. In this sense, it is different from the California 
system, which establishes a competitive market for providers of 
medical services. Part of the reason for these differences is 
that the ownership of hospitals is, on the whole, very different 
in Massachusetts, where the majority of institutions are 
voluntary, than in California, where they are generally investor- 
owned. It is easier for California to view a hospital as a 
business enterprise, while in Massachusetts, hospitals are 
regarded much more as an essential public service — similar to a 
public utility — and therefore regulated more heavily. 

Another important aspect of Chapter 372 is that because the 
business /industry coalition was responsible for developing 
it, they are now strongly supportive of the initiative. In 
California, business and the hospital industry did not 
significantly contribute to the shaping of the system, and in 
some instances, actively opposed it. This would be expected to 
make successful implementation much more difficult, and to make 
legislation much more controversial, which is what indeed has 
happened. 

The Massachusetts system varies from the Medicare DRG system 
in that it is an all payer system. It also limits utilization 
levels as well as reimbursement. 



40 



One aspect of Chapter 372 that has been controversial is the 
treatment of capital versus operating costs. Capital expenditure 
is virtually unregulated, but operating costs, such as labor, are 
limited. This has the effect of encouraging the purchase of 
machines which replace hospital workers, even when they increase 
the total cost of care. The state health planning agency has 
been deluged with requests for new capital projects, but some 
institutions are laying off hospital workers even as they 
purchase new equipment and build new buildings. It is still too 
early to measure the effects of the capital costs provisions on 
quality of care (e.g., will the infusion of new technology offset 
the the effects of possible staff shortages?) 

. Although policy analysts around the country are extremely 
interested in the results of Massachusetts' experiment, it is 
too early to judge the overall effects of Chapter 372. Many feel 
that the new law will "squeeze the fat" out of the system, 
resulting in cost containment of medical costs: others are 
certain that costs will simply be shifted, possibly to free- 
standing clinics and doctor's offices. This program will be an 
important one to monitor and compare to the other major models of 
prospective payment. , . . ._,a 

MICHIGAN 

As early as 1980, the State of Michigan had already made the 
transition from cost-based reimbursement to a prospective system. 
However, dissatisfaction with the limited effects of that system 
on cost containment led to the formation in early 1984 of a task 
force, appointed by the Health Department, to examine Medicaid 

41 



.■?',«•<• 



options. The task force looked carefully at the California model 
of competitive bidding and decided that that was not a viable 
option for Michigan. Rate-setting was also considered and 
discarded, for political reasons. The task force did agree, 
however, that a DRG system was a viable option, and a system 
similar to federal Medicare, but with Michigan-specific weights, 
was scheduled to be implemented statewide on January 1, 1985. 

Under the pre-1985 system, hospital revenue was based on a 
prospective budget estimate for each hospital that was developed 
from historical figures. Facilities were not reimbursed if they 
went over budget, but received incentive payments if they were 
under budget. This system covered only in-patient hospital care, 
although a similar system existed to cover nursing home care. 
Approximately 70% of total Medicaid expenditure in the state was 
covered by this type of PPS. Utilization controls included the 
surgical second opinion program and review of ambulatory surgical 
procedures . 

State officials felt that, although Michigan's system had 
prosective features, it was basically retrospective. This is 
because it relied on actual historical expenditures. They felt 
that the system did not offer the incentives to contain costs 
that were present in a DRG-based system. 

The system that was to begin in 1985 is based largely on 
Medicare DRG. It would use the same classifications, but with 
different weights developed specifically for Michigan Medicaid. 
The formulas would also be similar to Medicare's, except for the 
treatment of capital, which would be apportioned as a fixed 



42 



amount per case, rather than passed through, as in Medicare. 
Direct teaching costs would also be apportioned per case. There 
would be no peer grouping of facilities, but that would be an 
option for the future. Unlike Medicare, childrens' hospitals 
would be included, and the utilization controls of the old system 
would be retained and expanded. 

Psychiatric hospitals and distinct part units had been sche- 
duled to be included in this DRG system, but the reaction of the 
mental health community led to the exclusion of these services. 
They would instead be covered under a system of prospectively set 
per diem rates. This change led to postponing implementation of 
the system from October 1, 1984, to January 1, 1985. 

OHIO ,» - • . 

On October 1, 1984, the State of Ohio implemented a system 
based on peer groupings and a reimbursement formula methodology 
similar to Medicare's. The system covers Medicaid in-patient 
hospital care only, accounting for approximately 40% of the state 
Medicaid budget. 

Formerly, the state operated on a cost-reimbursement system, 
but cost containment considerations led to a search for an alter- 
native system. The recommendation to adopt the new system came 
from the Governor's Commission on Ohio Health Care Cost Contain- 
ment, convened in May, 1983. Development of the new system began 
in January, 1984. One prominent consideration was to make the 
system similar to the Medicare system. in this way, the hospi- 
tals could avoid making two radically different transitions in 
their operations. 



43 



The State Department of Public Welfare administers the 
payment formula, which uses the same DRG classifications as the 
Medicare system, but with different relative weights based on 
Ohio-specific Medicaid claims data. Approximately 60,000 records 
were used to develop these state-specific weights. The formula 
uses three basic components to calculate the payment amount: 

• average cost per discharge (by peer group); 

• a capital component; and 

• a teaching component (for hospitals with teachina 
programs ) . ^ 

The average cost per discharge is calculated for a peer 
group of hospitals. This system is intended to account for the 
costs of both different and comparable service settings. The 
groups are based on the wage index of the Standard Metropolitan 
Statistical Area (SMSA) where each hospital is located, and they 
were designed to permit the state to pay similar prices to 
hospitals with similar costs. Separate peer groups have been 
established for the one teaching hospital in the state, rural 
hospitals over 100 beds, rural hospitals under 100 beds, and 
childrens' hospitals. In this, last peer group, a common rate 
structure is not calculated. Instead, a facility-specific rate 
is formulated separately for each children's hospital. The peer 
group methodology was deliberately kept simple so that it would 
be easy to understand the divisions. 

The capital component is based on the percentage ratio of 
capital to total expenditure for the facility. For hospitals 
that are not teaching hospitals but that have teaching programs, 
two factors for direct and indirect teaching expenses are added 



44 



to the formula. These factors are based on the ratio of these 
expenses to in-patient care costs. 

Distinct part units, such as psychiatric and rehabilitation 
services, are included in the Ohio system. In contrast, Medicare 
excludes them from prospective payment. Freestanding rehabilita- 
tion facilities, long-term care hospitals, and SNF/ICF units are 
excluded under the Ohio system. Providers may keep the differ- 
ence between payment and cost if they come in under the DRG rate, 
and high outliers are paid at a percentage of the normal DRG 
rate. --• "■ ' 

A cap of 7.5% has been placed on inflation and expected rate 
of growth. A special facility-by-facility adjustment is also 
available for those serving "disproportionate numbers." 

The system is being phased in over three years by blending 
hospital-specific rates with peer group prices in declining pro- 
portions, 50/50 for FY 1985 and 25/75 for FY 1986. In this way, 
by July 1, 1986, it is expected that. all hospitals will be paid a 
price per discharge based on average cost of the peer group. The 
exception will be children's hospitals, which will continue to be 
reimbursed on a facility-specific basis. 

The state publishes a quarterly statistical report for the 
purpose of admission pattern monitoring. Readmissions , outliers, 
and transfers will also be carefully reviewed to assure 
appropriateness and quality of care. The regulations call for 
administrative review of each facility "not less" than semi- 
annually, for most aspects of PPS. "' 

The regulations also call for a commission to be convened by 
the Director of the Department of Human Services by November 1, 

45 



1984, to recommend any adjustments needed in PPS to recognize 
changes in technology. This commission is permanent and must 
report by March 1st of each year. 

Of course, it is difficult to predict the future development 
of a system that has only been implemented recently, but state 
officials feel there will be a good deal of interest in the 
inclusion of pediatric diagnoses in the system. This is one of 
the large controversies in the field of PPS and many other states 
will be interested in what happens in Ohio. 

OKLAHOMA 

The State of Oklahoma has experimented with various forms of 
prospective reimbursement systems since 1977, but the effective 
date of the present system is April, 1983. This system consists 
of an all-inclusive per diem rate for in-patient Medicaid 
services, calculated from cost reports and case-mix indices from 
a base year, adjusted by an updating factor. 

Hospital-specific per diem rates are calculated using the 
Medicare cost report for each individual hospital for the most 
recent cost reporting period ending on or before June 30, 1982 or 
June 30, 1981, whichever results in lower rates. These rates are 
then adjusted for estimated inflation between the end of the base 
period through January 1 of the cxirrent rate year. Adjustments 
to cost reports as a result of subsequent audits will not be 
considered in the current per diem rate for the calendar year. 
However, cost adjustments will be allowed for subsequent calendar 
years under the existing methodologies. The estimated inflation 
adjustment is based on the HCFA Market Basket Index, including 

46 



technology adjustments, as published most recently prior to the 
beginning of the rate year in question. 

Several additional conditions have been set under this 
system to control utilization and adjust for case mix. The 
maximxjm reimbursable length of stay is ten days and no 
readmissions are permitted within 20 days of discharge (with 
exceptions on a case-by-case basis ) . Where a determination is 
made by the Professional Standards Review Organization that a 
hospital patient is receiving services at an inappropriate level 
of care, payment will be made at the statewide SNF rate as 
established in the Medicaid State Plan. The prospective rate is 
subject to adjustment where a hospital's case mix in its rate 
year differs significantly from the mix in its base period. 
Where such a difference is demonstrated, the hospital's rate will 
be determined by recalculating its base period cost as if the 
hospital had experienced in its base period the case mix that it 
experienced in its rate year. 

Out-of-state hospitals for which base period costs are not 
available will be paid the lower of billed charges or the average 
prospective rate for all Oklahoma hospitals. Out of state 
hospitals which routinely provide services to Oklahoma Medicaid 
recipients will be required to submit cost information sufficient 
,to establish a prospective rate under these methods and 
standards. New hospitals for which there is no base period will 
be paid a prospective rate established from an average of 
prospective rates paid to existing hospitals of a similar type. 
Consideration will be given to number of beds, levels of care 



47 



provided, and geographical location. 

Preliminary reports indicate that, since the implementation 
of the prospective system, there has been no increase in in- 
patient state Medicaid costs. This is a significant development 
particularly in light of the 40 percent increase in Medicaid 
costs in the three years prior to system implementation. 
According to state officials, hospitals have been generally 
cooperative during the implementation of the system. Regional 
officials feel that the new Medicaid payment system, together 
with Medicare DRG, has effectively changed the philosophy in 
hospitals: they are becoming increasingly competitive and react- 
ing more directly to the marketplace. These officials also feel 
that the new system has had no detrimental effect on the quality 
of care in the state. 

PENNSYLVANIA 

PPS in the state of Pennsylvania consists of a DRG-based 
system with peer groupings. The classification system of 
Medicare is used, but with Pennsylvania- specific weights. These 
relative weights are to be redetermined annually by the state. 
Psychiatric, drug, and alcohol rehabilitation units of general 
hospitals that are enrolled in the Medical Assistance Program are 
excluded from the prospective payment system, as are children's 
hospitals. 

Costs that are excluded from the PPS rate and are paid 
separately include capital depreciation and interest, direct 
medical education costs, and costs for outliers. Readmission 
within seven days does not qualify as a separate admission, but 



48 



is included as a continuation of the original admission. 

The peer grouping system is complex. A chart illustrating 
the methodology appears in Exhibit 2.4. The system starts with 
four ranking systems, called "concepts." The four concepts are 
teaching status. Medical Assistance volume, environmental charac- 
teristics, and hospital costs. All hospitals in the state under 
the PPS program are ranked in order according to a number of 
criteria under each of these concepts (see (a) in Exhibit 2.4). 
For example, under the concept of teaching status, the criteria 
are the number of residency programs at the hospital, the number 
of full-time equivalent (FTE) residents at the hospital, and the 
ratio of FTE residents to total hospital beds. The rank numbers 
for these criteria are then added up; this number becomes the 
concept score for that hospital (b). For example, if a given 
hospital ranked 30th in the state on the niimber of residency 
programs, 35th in the state in number of FTE residents, and 25th 
in the state in the ratio of FTE residents to total hospital 
beds, that hospital would have a teaching status concept score of 
90. ■ - 

. For each concept, the hospitals are then divided into eight 
groups, with an equal number of hospitals in each category — 
four groups below the median and four above the median ( c ) . This 
creates 32 groups, called "concept categories." 

To determine the composite per diem rate for each hospital, 
the average per diem rate of all the hospitals in each concept 
category is computed (d). In this way, each hospital ends up 
with four "concept" per diem rates, the average of each concept 
group that particular hospital finds itself in. These four 

49 



(s) 



HospllaU'Ronktdby 
Concspl One, Crllarlon 1 


Hoipllol X = r,-,.,^^^ 
Hotpllal Vsr, ^ — 
Hospital z s r, 



(b) 



Hospllalt Ranked bu 
Concept One. Criterion 2 



Concept One Score 
Hoipttal X 



Hospitals Ranked by 
Concept One. Criterion N 






Four Concept ^ 
Categories '■* 



Four Concept 
Categories '■* 



i 




Hospital X z 



Hospitals Ranked Into 
Concept Categories by 
Concept One Score 



EXHIBIT 2. 4 PEER GROUPING riETIIOOOt OGV FOR 
HOSPITAL X. PENNSVLVANIA 



Concept Two Score 
Hospital X 



Concept Three 
Score. Hospital X 



Concept Four 
Score, Hospital X 



Same nelhodology 



(I) 



Hospitals Ranked into 
Concept Categories by 
Concept Two. Three, 
and Four Scores 



Average per diem Is cal- 
culated lor Concept Two. 
Three, and Four categories 
ol Hospital X 



FINAL PEER 
GROUPINGS 



(HI 



n 



Average par diem 
Is calculated lor all 
hospitals In this 
concept category 



(•) 



' —- Median 




(Per Olams Concept One • Concept Two > 
Concept Three • Concept Four): 

(Sum or Concept Per Olems) -^ 4 = 

Composite Per Diem Rate (or Hospital X- 



Four 
peer K 
groups 



lledlen-H 



Hospitals Ranked by 
Composite Per Olem 



Hospital X Composite 
Per Olem 



fFour 
peer 
groups 



(c) 



Notes 
Hospitals' means all providers In PPS program 
Each concept category contains an equal number 
of hospitals 

Hospitals can be In very dlllereni categories 
under each concept 



concept rates are then added up and divided by four, and this 
final figure is the composite per diem rate for the hospital (e). 
The hospitals are then ranked by composite per diem rate, and 
this ranking is then divided into eight equal groups, four above 
the median and four below. These constitute the final peer 
groups ( f ) . 

Pennsylvania law provides for a first-year phase-in based on 
75 percent of the hospital's average cost per discharge and 25 
percent of the peer group's average cost per discharge. In the 
second year, this shifts to 50/50. Each hospital is required to 
set up a utilization review process that must review every Medi- 
cal Assistance patient's need for inpatient hospital services. 
This includes a utilization review committee and a hospital 
utilization review plan. The state- level Department then moni- 
tors these individual facility programs. 

The utilization review committee must conduct admissions 
reviews, stay reviews for potential outliers, and medical care 
evaluation studies. Procedures for dealing with these areas must 
be outlined in the hospital's utilization review plan. Admission 
reviews must include the initiation of discharge planning to 
provide timely placement in an appropriate level of care for 
those patients that may require post-hospital care. Medical care 
evaluation (MCE) studies must identify and analyze medical or 
administrative factors related to patient care rendered in the 
hospital and, when indicated, make recommendations for changes 
that would be beneficial to patients, staff, the hospital, and 
the community. At least one MCE study must be in progress at any 



50 



time, and at least one must be completed each calendar year. 
UTAH 

Utah's Medicaid program is the oldest DRG-based system on 
the state level. Under federal waiver, a system was implemented 
in July, 1983, that allowed the state to "piggyback" onto the 
federal system, thus implementing a DRG-based prospective payment 
system almost immediately. 

Without phase-ins, complicated adjustments, peer groupings, 
state-specific weights, negotiated rates, or any of the other 
complex individualized administrative aspects of other state 
systems, the Utah system is comparatively simple and straightfor- 
ward. Since it is so much like the federal Medicare system, it 
needs little explanation. Since it is modeled on Medicare, it 
was also easy to implement quickly. This allowed Utah to have a 
PPS program up and running before everyone else. it allowed the 
state, in some ways, the best vantage point in terms of PPS 
development; if DRG-based systems do prove to be the wave of the 
future, Utah has a functioning system that is already saving 
money and that can be easily adjusted in light of future develop- 
ments. If the state decides that other options are more suit- 
able, it will already have a database to use in developing future 
options. 

Thus, Utah has adopted for its Medicaid program essentially 
the same DRG-based system as the federal government uses in 
Medicare. This includes the classification system for diagnoses 
and the weights based on resource use employed in the formula, as 
well as the pass-throughs. The rates of payment calculated with 



51. 



the DRG formula are reduced 20 percent across the board by the 
state. However, the state does not use this figure generated by 
the DRG formula as the amount of payment, but as a target rate in 
a discharge discount system. s.-rr j* - 

Under the discharge discount system, the DRG rate as 
calculated under Medicare becomes a target rate. Actual payments 
are then calculated within a "corridor" around this target rate. 
Providers receive payments of not more than 110 percent but no 
less than 60 percent of their costs. This is illustrated in 
Exhibit 2.5 for a theoretical discharge with a target (DRG) rate 
of SIOOO. If the cost to the provider is less than the target 
rate, the state pays cost plus 10 percent, but no more than 
$1000. For costs equal to or above the target rate, the state 
pays no more than $1000 unless the provider's costs exceed 
$1,666, i.e., until 60 percent of provider cost exceeds the 
target rate. From that point, the state pays 60 percent of provider 
cost. 

Prior authorization and concurrent review already existed in 
the state, and there are no further utilization controls built 
into the system. The rates are set by the state Medicaid agency, 
and they cover only Medicaid services. , >. . 

Once this system has been in place for a while, the state 
wants to develop an experience-based state-specific pricing 
structure, based on information collected during this period. 
The state is also studying the possibility of moving eventually 
to a selective provider contracting program. . 

WASHINGTON " 



52 



EIHIBIT 2.5 DISCHARGE DISCOUNT STSTEM 



State pays 
cost plus 10^ 



*\\\\\\\X\X\\\NXX\NX\\XN . 



fc\\X\XXX\XXXXXXX\XXXXX\X 



* X X X X X X 



k X X X X X X vrilTA rsa^C xxxxxxxx . 

fcXXXXXX * ' xxxxxxxx> 

y>x/yy . . . ••/•/•••. 

k X X X X X X fO^CTAr f"0 r A xxxxxxxx ■ 

»XXXXXX O XXXXXXXX" 



yyy/y/. yy/y/yyyy/yyyyyyyy 



X X X X X X 



y • • y • y 



• ••••••• 



fcXXXXXXXXXXXXXXXXXXXXXXX 
iXXXXXXXXXXXXXXXXXXXXXXX 

xxxxxxxxxxxxxxxxxxxxxxx 

fcXXXXXXXXXXXXXXXXXXXXXXX 

yy*^ ' 

« X X X 



^ ^**^**^^^*^j^^^**f*f 



xxxxxxxxxxxxxxxxx %x 
xxxxxxxxxxxxxxxxxxxxxxx 

XXX* XXXXXXXXXXXXXXXXXXX 

xxxxxxxxxxxxxxxxxxxxxxx 
kXXX *xxxxxxxxxxxxxxxxxxxh**- 



state pays 
60S of cost 



$0 



300 



1000 



300 



IF 

2000 



500 



3000 



Target (DRG) Rate 



PROVIDER COSTS 



52a 



The State of Washington used to operate as a rate-setting 
state, with rates set through budget review by the Washington 
State Hospital Coiiiinission, a representative body. In October, 
1984, however, the Commission's Task Force on Alternative Reim- 
bursement was in the final stages of developing an all payer DRG 
system, to be presented on November 15, 1984. 

The rate-setting system, which was mandatory for all non- 
federal hospitals, operated through a review of the cost centers 
in each hospital's annual budget. Budget screens were developed 
based on data for all the hospitals in the state. Costs which 
exceeded these screens were disallowed by the state, unless the 
provider was able to justify them. The budgets were also com- 
pared to those of similar providers in peer groups based on size, 
teaching level, case-mix, geographic location, and other vari- 
ables. 

A hospital-specific component for capital was also added to 

the budget. It was then reviewed a final time to assure that 

rates were reasonably related to costs. _ ,. 

The final set of rates 

developed from this process was then subject to public hearing 
before implementation. 

In late 1984, the state was developing a prospective payment 
system based primarily on Medicare. This would be an all payer 
system, scheduled for presentation on November 15, 1984. This 
new system would be very similar to Medicare DRG; the same 
classification system would be used, as well as the federally- 
developed weights and the HCFA case-mix index. The average cost 
per discharge would be separated into three parts, adjustable by 
different factors: the operating cost component, capital cost 

53 



component, and teaching costs component. 

The operating cost component would be derived by calculating 
the costs attributable to Medicaid services and dividing by the 
number of Medicaid discharges. This operating cost per discharge 
would then be a justed for the hospital's case-mix index and 
inflated to the payment period with the HCFA Market Basket Index 
plus one percent per year allowance for changes in technology. 

Both direct teaching costs and working capital (interest, 
depreciation, and leases) were scheduled to be included in the 
Washington system as adjusted costs per discharge, based on a 
standardized cost per discharge. 

The costs of teaching and capital would be divided by the 
total patient days of the hospital. This per diem capital cost 
would then be multiplied by the number of Medicaid patient days 
at that facility, and divided by the number of Medicaid dis- 
charges. This would yield the capital or teaching cost per 
Medicaid case. With some adjustment to update these figures to 
the current year, these factors would be the total reimbursement 
in these areas. There would be no pass-through or adjustment for 
building costs, and the indirect costs of teaching would be 
included in the DRG methodology. 

Outliers would be treated in a similar manner as in the 
federal program: by setting aside a percentage of the budget to 
cover the projected number of cases that can be expected to fall 
outside the parameters established for that DRG. Also, there 
would be no adjustment for "disproportionate numbers," since it 
was felt that, in this state, as well as in Medicare, circum- 



54 



stances such as these were already taken into consideration in 
the methodology, and no special adjustments were needed. 



DOCUMENTS REVIEWED FOR DETAILED DESCRIPTION OF 
TEN STATE MEDICAID PPS 

General (information on more them one state or on PPS in 
general ) 

Report on Hospital Prospective Payment Systems Mandated by 
Section 2173 of Public Law 97-35 — Health Care Finance 
Administration. 

Medicaid Program Evaluation Cluster II: In-Patient Hospital 
Reimbursement — Abt Associates. 

A Catalogue of State Medicaid Program Changes — National 
Governor ' s Association. 

State Prospective Payment Systems: New Directions in Hospital 
Cost Containment — Intergovernmental Health Policy 
Project. 

DRGs and the Medicaid Program — Intergovernmental Health Policy 
Project. 

State Efforts at Health Care Cost Containment — National 
Conference of State Legislatures. ' " 

Diagnosis -Related Groups: The Effect in New Jersey, the 
Potential for the Nation — Health Care Finance 
Administration. 

California 

Competition in the Health-Care Marketplace, — New England Journal 
of Medicine, March 31, 1983 

Progress Report on Hospital Cost Control in California — New 
England Journal of Medicine, July 28, 1983 

Massachusetts ■ • . .' ^ >. £" j...-^n 

Terms of Endowment: Prospective Hospital Reimbursement in 
Massachusetts — Health Policy Advisory Center Bulletin. 
Vol. 15, No. 2. 

What Price Cost Control? Massachusetts' New Hospital Payment 
Law — NEJM, March 3, 1983. 

55 



Massachusetts and California: Two Kinds of Hospital Cost 
Containment, NEJM, March 3, 1983. nospiTiai cost 

A 3^=^5";j^^ Paper on Chapter 372: The Massachusetts Cost 
Control Law — Massachusetts Hospital Association. 

^"'"'"Associftio^'"'" °" '•''' -Massachusetts Hospital 

Michigan 

Medical Assistance Bulletins: 

l5^fiS"«l'nfl 2^° Prospective Reijnbursement Program 
I^^«n"f^2f' Diagnosis Relate Groups (DRG) 
#5360-84-09: Out-patient Fee Increase and Indigent 

Care Adjuster 

I53IS'«a"^?*' ?JJ:^f^^®''^ Substance Abuse Services 
#5360-84-11: (Utilization Control) 
#5310.1-84-12: (Utilization Control) 

Ohio 

Draft State PPS Regulations 

Issue Papers on PPS 

Pennsylvania 

Draft State PPS Regulations 

Utah 

SPCP Proposal by Pace Associates 

Washington 

Proposed State Medicaid Plan (not yet approved as of October 



56 



'.hi. 



■U' 



CHAPTER 3 

THE POTENTIAL MEDICAID CONSEQUENCES OF 
HOSPITAL RESPONSE TO MEDICARE DRG 

Two tracks of Medicare DRG impact on Medicaid were described 
in Chapter 1. The first follows Medicare DRG as a stimulus to 
change in Medicaid in-patient reimbursement, the subject of 
Chapter 2. The subject of this chapter is the track that follows 
Medicare DRG as a stimulus to hospital and local delivery system 
change that, in turn, impacts on Medicaid patients, other 
providers, and /or program expenditures. 

To identify potential DRG spillover impacts onto Medicaid, 
and to begin constructing an effective evaluation approach to 
this issue, a series of exploratory interviews was initiated with 
hospital administrators and health-care analysts. Interviewees 
were selected to represent a range of different perspectives. 
They included local hospital financial administrators, 
professional association staff, and policy analysts - all 
familiar with Medicaid and working in the areas of health finance 
and service delivery. Most interviews lasted approximately two 
hours and were loosely structured around the following topics: 
o expected hospital responses to DRG; 
o effects of those responses on Medicaid patients; 
° and^""^^ on other providers (hospitals and physicians); 
o effects on costs to the Medicaid program. 
A taped transcript of a session on spillover effects, from a 
conference conducted by HCFA and the State of New Jersey, was 
also reviewed. The issues emphasized in that session were 

57 



integrated into the major findings of this chapter. 

3.1 Medicare DRG Is Expected to Affect Medicaid, and Impacts Are 
Likely to be Mixed 

The major finding that emerged from these exploratory 
discussions was that the Medicare DRG program is likely to have 
substantial spillover effects onto Medicaid within the hospital 
setting and across hospitals operating interdependently in health 
care systems. In fact, while Medicare DRG began only in October, 
1983, and was to be phased in gradually over a three-year period, 
most respondents agreed that spillover effects on Medicaid were 
already occurring in 1984 and that these effects needed to be 
examined soon in greater detail. 

A sxonimary of potential hospital responses to DRG and their 
effects is presented in Exhibit 3.1. Possible hospital responses 
have been grouped into five general strategic areas, each of 
which, in turn, includes one or more different courses of action 
that might be taken. In direct response to DRG, hospitals are 
expected to devote a great deal of attention to (1) reducing 
average length of stay — overall and in problem DRG categories, 
and (2) improving the efficiency with which services are 
delivered. These two strategies are likely to be implemented 
hospital wide, thus affecting Medicaid and other patients as well 
as Medicare patients themselves. Consequently, net savings in 
service delivery and negative effects on patient outcomes, if 
any, will presxomably extend to the Medicaid program. 

Some hospitals may attempt to shift Medicare costs to other 
payers, including Medicaid (3), or "diimp" financially undesirable 



58 



EXHIBIT 3,1: 



SUMMARY OF POTENTIAL MEDICAID CONSEQUENCES OP HOSPITAL RESPONSE TO MEDICARE DRG 



Pottntial effects 



NoapLCAl tt«spona« to oac 

l.HotplCAll 4dOpC poUCX*! 

•nd practices to rvdueo 

Icnqcn o( stay (LOS) for 

«il patients, at laast 

in certain diaqnostic 

cataqories: 

o Carltar/nore frequent 

diichacqe to ho«M care 

or lon^-teni care, or 

transfer to out' 

patient services 
o Hore use of out*patient 

for diaqnoktic and pre- 

surgical teatmq 
o Increased preasurs and 

incentives for physi- 

ciana to reduce LOS. 



On Medicaid Providers 



On Medicaid Patients 

Negative: 

o Premature discharges for 
transfers to out-patient 
services could lead to 
increased readsiiasions. 
increased probability of 
developawnt of cnronic 
conditions, ate. 



On Proqrea 
Costs/Adsiinistratioe 

Positive 1 

o Peduced LOS could result i 
savings on in-nospital 
costs for Medicaid 
patients. 
Negative: 

o The negative affects of 
preswture diactiarges on 
patients could lead ulti- 
• aately to increased costs 
for Medicaid. 



-Hespitala adept new ser- 
vice deUv«ry afficiency 
Measures for all patients 
includingi 

o Peduced use of ancil- 
lary Medical services 
o Reduction/elioination 
of certain non-nedicai 
lervicea and aawnities 
o Increaaed uae of out- 
patient care in place 
of in-patient care 
o Increaaed Monitoring 
and peer review of phy- 
sicians service 
util isation 
o Oevelopatent of eore 
tiMely. accurate racord 
keeping, and billing 
procedures 
o Development of alter* 
native service options, 
such as hospital -based 
long-term care 
o Collaboration vith 
other providers to 
share expensive services. 



Hogstives 

o If service reductiene 
are too SKtonsive. qual- 
ity of patient care 
could be reduced. 

o tf out-patient is inap- 
propriately used to re- 
place in*p«tient care. 
quality of care could 
suffer. 



Positive! 

o Mew efficiencies could lead 
to reduced costs to the 
Medicaid program. 
o To the estent tlut better 
record and billing systeaw 
reap savings that outweigh 
tAe&r costs, overall costs 
for serving Medicaid 
patients will decrease. 

Negative: 

o To the extent that hospi- 
tals proawte additional uso 
of out-patient services to 
cover m-patient Looses. 
Medicaid coats could 
increaae. 



3.Hoapitals shift costs of 
Medicare patients onto 
non-nedicare patients, 
including Medicaid. 



.Private hospitals 'diMp' 
financially diaadvanta- 
^aoua patients (Medicare 
and Nedicaidt into public 
nospitala through trans- 
fers or strict admiaaion 
policies. 



Hegativoi 

o To the estent that Medicaid 
staff fail to identify and 
prevent cost-ahif ting. Med- 
icaid coats could increase, 
eapecialty in retrospective 
cost-reiaiourssme n t systems. 



Negative t 

o Public hospitals could 
suffer additional 
losaes since they can- 
not turn away these 
patienta. Some nay be 
forced out of business 



Negativet 

o Medicaid and Medicare 
patients could become 
more concentrated in 
public hoapitala. where 
the risk of low«r qual- 
by those hespitala fin- 
ancial conatraints. 
o both "dumping* and the 
closing of some public 
hospitals could result 
in reduced access or 
choice for Medicaid 
patients. 



Negative I 

o Medicaid programa may be 
forced to provide addition- 
al rate adjuatmenta for 
public hospitals, thus in- 
craaaing overall Medicaid 



.Hospital a alter case eix 

to include: 

o Mora paying patients 
and fewac Hedicara/Mod- 
icaid patients 

o Here patienta in pro- 
fitable ORG Classifi- 
cations 

Mors caaes in areaa in 
which they are aape- 
ially cost-efficient. 



Negative: 

o Public hoapitala could 
be forced to care for 
mora Medicare/Hedicaid 
patients and more pat- 
ients in financially 
disadvantageoua ORG 
classif icationa. 
Their financial situa- 
tiona could deterior- 
ste aa a result. 

o Public hospitals could 
be forced to taka nore 
cases in the lesa «<- 
tieient service areas. 



Negativet 

o Some service areas may 
be undersuppl led. leav- 
ing patienta to cope 
with less access than 
before. 



Positive/Negative I 

o In sosM areas, service 

afficiency could increaae 
and costs could decreaae. 
In others, the opposite 
night occur. 



58a 



Medicare, Medicaid, and other patients into the care of public 
hospitals ( 4 ) . These two strategies , however adaptive for the 
hospitals that use them, are likely to have negative consequences 
for Medicaid patients, public hospitals (a major source of 
Medicaid services) and program costs. Finally, hospitals may 
choose to manipulate their case mixes in positive ways by 
promoting admissions and services in areas that bring a higher 
return on costs (5). The potential effects of this strategy are 
not encouraging for public hospitals and Medicaid patients, but 
where hospitals can cut costs through efficient specialization, 
some savings for the program could be realized. These five 
strategies and potential effects are explored in more detail in 
the remaining sections of this chapter. 

It should be noted that hospitals are expected to vary 
widely in degree of need to change in response to DRG and in how 
they respond to DRG. Size of Medicare caseload is an obviously 
important factor, as well as is the pre-DRG general degree of 
efficiency with which Medicare and other patients are served. As 
for the strategies they adopt to cope with Medicare DRG, some may 
opt for concentration on LOS reduction, others on administrative 
or service delivery efficiencies, still others on case mix ad- 
justments or cost shifting, and so on. It is likely that various 
combinations of these strategies will be adopted and will be more 
or less successful (either from the hospitals' point of view or 
that of the Medicaid program), depending on several factors. As 
outlined in Exhibit 3.2, some of the more important of these 
factors are the following: whether the hospital is privately or 
publicly owned (the former having more flexibility in whom they 

59 



EXHIBIT 3 2: FACTORS LIKHLT TO AFFECT HOSPITAL ADAPTATIOH TO DRG 



Medicare 
DRG 



m 



Some hospital3 experience a 
greater need to change 
policies, procedures, service 
strategies than other 
hospitals. 



1 



High vs low Medicare 
caseload. 

Less vs. more pre -DRG 
efficiency in treating Medi- 
care patients, particularly 
in high volume DRG 
categories. 



Some hospitals are able to 
adapt more successfully 
(from their own point of 
view) to DRG than otl-ier 
hospitals. 



1 



• Private v. public ownership 

♦ Less V more stringent 
Medicaid & oU-ier third party 
reimbursement policies. 

♦ More vs. less sophisticated 
record and billing systems. 

♦ More vs less system -wide 
opportunity for specialization 
& collaboration. 

• Fewer vs. more constraints on 
development of long-term 
care & home health services, 
hospital -based and other. 

•More vs. less ability to shift 
loss -producing patients to 
otJier hospitals. 

• More vs. less ability to influ- 
ence physician admitting and 
treatment practices. 



serve, the latter being the service provider of last resort); the 
degree of reimbursement stringency and careful cost-monitoring 
exercised by Medicaid and other third party payees in the hospi- 
tal's area (allowing more or less opportunity for cost shifting; 
for example ) ; the degree of system-wide opportunity for efficient 
specialization or collaboration (small systems offering less 
opportunity than large ones ) ; and the degree to which local 
certificate of need regulations allow development of alternatives 
to prolonged inpatient hospital care. 

As can be seen from Exhibits 3.1 and 3.2, the large urban 
public and teaching hospitals are expected to be most affected by 
DRG and to show the most strongly marked spillover effects onto 
Medicaid. These are typically the institutions that serve the 
largest numbers of medical assistance patients and they usually 
serve large numbers from both programs side by side. These 
hospitals, however, generally know that they will be affected 
substantially, and since they tend to have fairly sophisticated 
staffs, they are preparing well to deal with DRG. In particular, 
they are most likely to have the in-house expertise, as well as 
the information system capabilities. In spite of their 
administrative capabilities, however, the large inner-city public 
hospitals have most to lose from such practices as "patient 
dumping" and specialization in profitable services among non- 
public hospitals. Public hospitals neither deny certain kinds of 
services nor service in general to Medicare or Medicaid patients. 

Just as the responses of hospitals to Medicare DRG are 
expected to be complex, the consequences for Medicaid may be very 



60 



mixed. Both positive and negative effects on Medicaid patients, 
providers and state/ federal payers are likely, and it is not 
clear how the program will fare overall. Further complicating 
the relationship is the growing movement among State Medicaid 
programs to implement various prospective payment systems of 
their own, which create additional or overlapping incentives for 
hospital behavior change. For ease of discussion in this 
chapter, several likely effects of Medicare DRG are separately 
outlined, but it should be remembered that each is part of a 
complex reaction in a service and payment system that defies 
simple description or prediction. 

3.2 Medicaid Costs Are Expected to Decrease from Reduced Length 
of Sta^r and Increased Efficiency, but at the Possi ble Risk 
of Negative Patient Outcomes 

Under DRG, providers are paid a predetermined fee for each 
patient discharged, depending on the patient's DRG 
classification. Each DRG is based on diagnosis, treatment 
procedures, age, sex, and discharge status; and payments are 
based on the allowable operating cost per discharge and other 
factors. In response to this fixed payment, it is expected that 
providers will develop patient treatment protocols and procedures 
which will limit both length of stay (LOS) and service related 
expenditures. For example, hospitals are expected to increase 
their use of outpatient care, institute monitoring of LOS and 
service levels, develop rigorous department level procedures for 
limiting ancillary services, and so on. 

It is anticipated that for various reasons, hospitals and 
other providers will not limit their efforts to Medicare 



61 



patients, but will extend their protocols and procedures to cover 
all patients, including those receiving care through the Medicaid 
program. These reasons include uniformity of administration and 
record-keeping, extension of cost-savings to other payers, 
avoidance of discrimination, etc. This, in turn, may drive down 
costs associated with serving Medicaid patients — representing a 
major positive effect of DRG on Medicaid. The efficiencies 
achieved in LOS and service delivery could have deleterious 
effects on patient outcomes, however. Less in-hospital service 
could mean lower quality care. These hospital responses and 
their potential impacts on Medicaid are summarized in Exhibit 3.3 
and in the remainder of this section. 

LOS. Various forces have caused hospitals to decrease LOS 
over the last few years, but the DRG system will exert increased 
pressure, since the hospitals will not be paid for the extra 
days. For example, one hospital administrator indicated that his 
institution has been able to decrease average LOS from over 
eleven days to less than nine. This is a process that started 
before the advent of DRG, but which was certainly spurred on by 
the new form of payment. This decrease in LOS has affected all 
types of patients ~ including Medicaid patients. Increased 
pressure has been applied to attending physicians to reduce LOS. 
One problem at this hospital, however, has been an inadequate 
supply of long-term care alternatives which necessitates extend- 
ing expensive acute care treatment for many who do not require 
such a high degree of care. Recently, the hospital has been 
promoting the increased availability of long-term beds, while 



62 



EXHIBIT 3.3: POSSIBLE EFFECT OF REDUCED LENGTH OF STAT AND OTHER EFFICIENCT MEASURES 



Medicare 
DRG 



Kl 



Hospitals change administrative and 
patient care policies and procedures 
to reduce average LOS, especially 
among loss -producing DRG cases: 

• Pressure /incentives are applied to 
physician behavior (peer reviev/, 
admitting privileges, etc.) 

• Outpatient services (for tests, post- 
surgical recovery, etc.) are used in 
place of in-patient services. 



Hospitals increase efficiency in 
services provided through: 

• Better detailed and more timely 
records of diagnoses and treatment. 

•Automation of data files. 

•Peer review of utilization patterns. 

• Elimination of non-essential 
services/procedures. 

• Collaborative resource use with 
other providers. 

•Development of alternative 
service delivery approaches. 





LOS and other cost 
elements are re- 
duced for Medicaid 
patients, especially 
where treatment 
overlaps with 
common Medicare 
categories. 






— ► 




Costs to the 
Medicaid 
program are 
reduced. 














— ► 


Premature 
discharges and 
transfers to 
reduce LOS, and 
reductions in 
services for the 
sake of efficiency 
could adversely 
affect patient 
progress and 
result in readmis- 
sions and develop- 
ment of chronic 
conditions. 








Cost to the 
Medicaid 
program are 
increased. 


f 




• 



other hospitals in the same market are starting or expanding in- 
home care programs. 

This interviewee noted that his institution is not offering 
financial incentives to staff physicians to decrease LOS. He 
felt that extension of admission privileges to attending 
physicians is a more powerful tool for gaining their cooperation 
in decreasing LOS. He also indicated that peer pressure will 
certainly be put on the "loss leaders," and as the physicians 
become more aware of the financial realities that drive the 
hospital, they will become more sensitive to the issue of 
uncompensated care. This interviewee noted that, in the future, 
if physician reimbursement is included in the DRG rate, a 
financial incentive to the physician to reduce LOS would be a 
possibility. 

A second hospital administrator also indicated that a 
primary effect of DRG on hospitals will be a reduced average 
length of stay. He anticipated that protocols and procedures 
adopted by hospitals to reduce length of stay for Medicare 
patients would "spillover" to treatment of Medicaid patients. 
This interviewee noted that LOS was a major problem for his 
hospital under the prospective payment system. Currently, his 
hospital's average LOS for Medicare patients (across all types of 
treatment) is ten days eUsove the overall average calculated under 
the DRG program — 17.5 days versus 7.5 days. This translates 
into a major revenue problem for the hospital, which has a high 
proportion of Medicare and Medicaid patients. He noted that 
recently his hospital had begun to institute policies to reduce 
LOS — policies which had already resulted in a drop of half a 

'63 



day on the average stay over the last seven months. He conceded, 
however, that this was only a beginning and that much more had to 
be done. 

The problem of decreasing LOS can be further disaggregated 
into effects on diagnoses that fall within the DRG classification 
system and effects on diagnoses that do not fall under DRG, or 
that are much less frequent under Medicare than under Medicaid. 
These include such cases as maternal /child health, pediatric 
diagnoses, and well-child care. For diagnoses liable to be 
treated under either program, Medicaid is Ixxely to benefit from 
LOS reductions due to the pressure of Medicare. However, in areas 
not extensively covered by DRG: 1) Will LOS decrease? and 2) If 
so, is this decrease due to the effects of DRG or to other 
factors? 

, .. Administrative Efficiency. Hospitals are not only expected 
to reduce LOS, but also to institute other administrative and 
service delivery efficiencies. For example, under the DRG system, 
more detailed and timely records of diagnoses and treatments will 
have to be maintained by attending physicians and hospitals. 
Interviewees indicated that any changes in documentation 
procedures and forms, as well as automation of data files, would 
"spillover" to Medicaid patients, as well as patients covered by 
other payers. It would be neither feasible nor efficient to 
maintain different reporting systems and documentation for 
different types of patients. Such improvements in record-keeping 
and data processing would provide the tools to facilitate both 
hospitals' identification of areas where other efficiency 

64 



measures are needed and likely to pay off, and routine monitoring 
of progress in such efforts. 

One administrator expected that it would be necessary to 
"educate" physicians to complete timely, detailed record-keeping 
on each patient. Under the DRG system, physicians are required 
to document and sign principal and secondary diagnoses within 72 
hours of the patient's discharge, which represents a radical 
change for many physicians. He also noted that his hospital has 
added three additional staff members to deal exclusively with 
medical records. These medical records technicians will support 
physicians in documenting diagnoses, and in making sure that all 
necessary information is available for siibmission to the 
intermediary. 

A second administrator said that his hospital was improving 
its record-keeping to make sure that proper coding and billing is 
occurring. This hospital already has a mainframe computer, but 
it has recently also purchased several IBM PCs specifically to 
run the DRG Grouper program. Hospital staff will be duplicating 
Medicare processes carefully in order to predict reimbursement 
for each discharge. They will then carefully compare actual 
reimbursement with expected, and check for errors, both on their 
part and on the part of the intermediary. In addition, this 
hospital has also put increased pressure on attending physicians 
to record diagnoses and treatments accurately, and also to "sign 
off" in a timely manner on discharges, as mandated by Medicare 
DRG. Reportedly, these changes have been implemented across the 
board, regardless of payer. ""'• ■^■.'■.v■ ^i:^ 



65 



These improvements in record-keeping and invoicing are 
likely to have a positive impact on overall administrative costs 
for hospitals, providing they do not cost more to implement than 
they are worth. To the extent that hospitals save money through 
better information systems, costs for Medicaid are likely to be 
reduced as well. 

Efficiency Through Reduced Use of Ancillary and Unnecessary 
Services and Increased Use of Outpatient Services. Under DRG, 
hospitals are expected to shift as much care as possible to the 
outpatient department. There, costs of services are lower, reim- 
bursement is still retrospective, and maintenance of costly acute 
care infrastructure can be reduced. Hospitals are expected to 
institute this change for all types of patients — including 
Medicaid patients. But such a change may or may not produce 
savings for the Medicaid program. As long as providers do not 
increase cost or volume of outpatient care to make up for losses 
in in-patient care, then this trend may well produce savings for 
Medicaid. it is expected, however, that shifts to out-patient 
care will occur. For example, one hospital administrator out- 
lined an aggressive program to increase volume of out-patient 
care, including contracting to provide routine examinations re- 
quired by law for city employees. This same hospital is also 
trying to increase use of the out-patient department by shifting 
patients recuperating from surgery to out-patient care rather 
than having them remain with in-patient departments. In addi- 
tion, since outpatient services, like emergency rooms, are tradi- 
tionally a major source of admissions, this hospital is trying to 



66 



lai 



develop an in/out surgery capability. In terms of spillover 
effects onto Medicaid, the interviewee emphasized that all pa- 
tients would be equally exposed to these new service approaches. 

Also, shifting the less critically ill to outpatient care 
will actually raise average LOS and resource use, since those who 
will still require acute inpatient care will be the more 
critically ill. In areas where Medicaid reimbursement is based on 
average LOS for large groups of hospitals, the hospital pursuing 
this strategy could ultimately suffer higher losses per patient. 

Finally, quality of care overall may suffer, since there may 
be a tendency to place patients in outpatient care who really 
could be better served through in-patient care, or to shift 
patients to an out-patient department before they are actually 
ready. This could easily happen, for example, in the hospital 
cited above that is trying to shift recuperating surgery patients 
to out-patient areas. Increases in readmissions , admissions to 
SNFs, and increased morbidity, could cancel out any savings based 
on shifts to outpatient care; in the extreme, they could even 
raise the costs to the Medicaid program. 

., Hospitals are also likely to decrease the use of expensive 
ancillary services to patients under Medicare DRG. It is unclear 
whether this will mean that hospitals will try to compensate for 
losses in ancillary revenues by increasing these services to 
other types of patients or will institute across the board 
changes in use of ancillary services for all types of patients. 
Other services, such as psychiatric and rehabilitative services, 
that are exempted from DRG, may see an increase in their 
populations, as physicians transfer patients from other 

67 



departments to these areas where the reintbursement is still cost 
based. For example, seriously-ill or dying patients may be moved 
to psychiatric wards once LOS is exceeded, on the basis that they 
are suffering from illness-related depression. 

Finally, there already is evidence that some hospitals are 
picking and choosing who will actually be admitted. This may be 
an effort to avoid incurring the cost of treating patients 
perceived as "loss leaders." For example, many providers are 
trying to screen out the "worried well" or those better off in 
long-term care or out-patient care. If present, this positive 
form of screening is almost certain to extend beyond Medicare 
patients to Medicaid and other payers. "Negative" screening, in 
forms such as patient dumping, is also being documented. This 
form of screening is discussed later in this chapter. 

Efficiency Through Inter-Hospital Collaboration and 
Development of Alternative Delivery Approaches, in response to 
DRG and other cost containment pressures, some hospitals are 
expected to band together in various kinds of collaborative 
resource sharing arrangements so that certain expensive services 
that can be shared are not duplicated across hospitals. Others 
may opt to make collaborative arrangements with providers of 
services that serve as alternatives to acute care when the latter 
is no longer necessary. For example, some patients would be more 
economically served if transferred to long-term care 
institutions, hospices, or home health services. Some hospitals 
are even developing their own alternative care service units to 
get around the external constraints on access to alternative 

68 



care. 

Efficiency through Closer Monitoring and Peer Review of 
Physicians . Associations and institutions report that the burden 
of changing practices is being placed squarely on the Medical 
Directors and department level staff. Several interviewees 
indicated that internal conmiittees would be formed to review the 
patient records of each physician. This would be done for all 
patients, including Medicaid patients. For example, one 
institution had already developed a series of reports on service 
utilization by physicians and by DRG. Patients were monitored 
from the point of admission through discharge for use of 
services, as well as for LOS. Peer review was not highly 
effective in the past, because it was not linked to the financial 
component; DRG should change that substantially. 

! ■ - ."■ 

• "' ■ Effects on Medicaid Patients . Most of the changes in 
hospital and physician behavior discussed above are expected to 
contribute to decreased costs for the Medicaid program. It is 
not very clear, however, what the impact of these changes will be 
on the quality of care for Medicaid patients. The discussion 
about out-patient care began to outline some of these very 
difficult problems. Possible negative consequences for patients 
include increased readmissions due to prematxire discharge for the 
sake of reduced LOS, increases in incorrect or delayed diagnosis 
due to less liberal use of diagnostic tests and procedures, 
increases in complications due to decreases in the overall atten- 
tion given by medical personnel per case, and so on. While there 
appears to be no evidence to date that patient care is suffering 

69 



r- 



at the expense of the DRG- inspired efficiency raeasiires described 
in this section, it is possible that, at some point in the fu- 
ture, some hospitals (perhaps those with particularly weak pro- 
fessional review and strong pressures to economize) will begin to 
sacrifice quality for the sake of economy. 

3 . 3 Medicaid Ma^ Suffer Cost Increases and Other Adverse Impac ts 
Due to Cost Shifting and "Patient Dumping^ 

Respondents generally agreed that DRG payment will not 
adequately cover in-patient care costs for many institutions. 
Two implications can be drawn from this. First, "self -pay" 
patients may be pursued more aggressively for payment, and 
hospitals will be looking for new ways to deal with bad debts. 
Second, it means that level of services and charges to non- 
Medicare patients may increase. This phenomenon, referred to by 
many as "cost shifting," and its potential impact on Medicaid are 
outlined in Exhibit 3.4. 

In response to hospital attempts to shift costs, it is 
expected that Medicaid and other payers will carefully monitor 
patterns of reimbursement to ensure that such manipulation does 
hot occur. One hospital administrator noted that since the 
implementation of the DRG program, Medicaid has become extremely 
aggressive about checking eligibility, auditing bills and cost 
reports, etc. For example, cost reports were very closely 
examined to determine whether hospitals had overcharged. 

Some state Medicaid programs may follow the lead of the 
Medicare program with respect to inflationary increases for 
hospitals. For example, one interviewee noted that, in the past. 



70 



EXHIBIT 3 4: THE POTENTIALLY NEGATIVE EFFECTS OF COST-SHIFTING 



MEDICARE 
DRG ^ 



o 



The extent to which Medicaid staff 
monitc>r patterns of service and 
charges will affect hospital ability 
to shift costs. 



Some Hospitals 
attempt to com- 
pensate for DRG 
losses by 

increasing services 
and charges for non- 
Medicare patients, 
including Medicaid. 



Costs to the Medicaid 
program are increased, 
especially in non-PPS 
states. 



The extent to which Medicaid 
programs independently put 
prospective and other limits 
on services and costs per 
service will affect hospital 
ability to shift costs. 



hospitals in this area have asked for, and received, an annual 
rate increase of approximately 10 percent for inflation under 
Medicaid. This year, however - while hospitals have asked for a 
10 percent increase — it appears that the state program will 
limit the increase to the rate increase used by the Medicare 
program (i.e., 6.9 percent minus 1 percent). 

Instead of (or in addition to) shifting costs, some private 
hospitals may attempt to avoid serving Medicare patients who are 
more likely to become Medicare loss -producers or outliers or 
eventual Medicaid clients. As outlined in Exhibit 3.5, private 
hospitals may, through pre-admission screening and transfer 
policies, "dump" economically risky patients into the public 
hospitals where they cannot be refused service. Such dumping 
could have very adverse effects on public hospitals' ability to 
survive, and on Medicaid patients' access to care and quality of 
care received. For example, one interviewee indicated that since 
the implementation of the DRG program, his hospital had seen a 
substantial increase in the number of transfers from other area 
hospitals. In particular, it was noted that many patients who 
could not pay for services at other hospitals were being sent to 
this public hospital. If transfers increase, public hospitals, 
which cannot turn anyone away, will end up carrying the other 
hospitals in the area, without increased compensation for 
performing that service. Some may go bankrupt in the process, 
thus producing an even greater service gap for Medicaid patients 
in the future. 

The more such transfering and skimming occurs, the more 
likely the medical care system will become a two-tiered one: 



71 



EIIIIBIT 3 5 NEGATIVE EFFECTS OF PATIENT DIJMPING 



MEDICARE 
DRG 



m 



Non -public bospiuis 
instjlule pre- 
admission screens 
designed to avoid 
admitting Medicare 
patients in loss-leading 
categories oi likely to 
exceed DRG LOS limits 










These screens, or 
oUier in addition. 




also (liter 
Medicaid 


' oui some 
patients 





Public hospitals, 
which cannot 
refuse patients, 
(ind vast increases 
in high cost and 
Medicaid cases 



Public hospitals 
suffer losses due to 
DRG constraints and 
low Medicaid rates 
(especially in PPS 
states ) 






i: 



Some public 
hospitals reduce 
service capability 
or close altogether 



Overall access to 
hospital services 
lor Medicaid 
patients is reduced 



Stales make 
additional 
adjustments in 
Medicaid rales 
lor public 
hospiuis 




Costs 10 the 
Medicaid program 
are increased 





private-based care for those who can pay, and over-burdened 
public care for those who cannot. 

3.4 Adoption of Other Case Mix Strategies by Hospitals May Lead 
to Mixed Results for Medicaid 

Transfering and screening out certain potentially expensive 
Medicare clients (discussed aibove) is one strategy hospitals 
might take to msUce their Medicare case mix more economically 
viable. The negative consequences to Medicaid of that strategy 
are quite apparent. Some hospitals are likely to try other case 
mix approaches in response to DRG that may have mixed impacts on 
the Medicaid program. These additional strategies are outlined 
in Exhibit 3.6. 

Among the more likely strategies hospitals may adopt are the 
following: 

(1) Increasing admissions and case loads — especially 
increasing the numbers of paying patients and increasing the 
caseloads of doctors who practice in a manner that maximizes 
reimbursement to the institution. For example, one of the major 
strategies one hospital plans to use to cope with financial 
impacts of prospective payment is to increase admissions. Until 
recently, only doctors directly employed by this particular 
institution had admitting privileges. In an effort to enlarge 
and diversify the patient population (in response to the DRG 
program) , this policy has recently been relaxed: independent 
practitioners are now being encouraged to admit their patients to 
the hospital. An additional incentive which the hospital hopes 
will increase admissions is lower rates than at other area 



72 



EIIIIBIT it: 



MlliO POTENTIAL EFFECTS OF HOSPITAL 
STRATEGIES TO CHANGE CASE MIX 



Medicare 
DRG 



M 
PI 






Hospitals adjust policies 
and s«rvic*s t^ Incrcas* 
caseload in flinnctally 
advantageous DRG 
categories and to 
■decrease loss-produang 
cases by: 



>Llt>e[alizlng physician 
access policy to allow 
efficient physicians and 
ttiose who specialize in 
desirat>le DRG areas to 
admit patients 

Eliminating loss-produc- 
ing service specialties & 
developing or eipanding 
profitable ones 



• Increasing overall 
volume by Increasing 
admissions 

■ Specializing In services 
Uiey perform particu- 
larly efficientiy. 



Across hospitals 
in a service area, 
tiie supply of non- 
lucrative. yet 
necessary services 
could decrease 



Some service gaps 
could result In re- 
duced access (and 
higher costs) (or 
Medicaid as well 
as other patients 



Public hospitals 
could be (orced to 
meet demand (or 
non -lucrative 
services 



Overall costs o( 
tome hospital 
services could 
decrease (or 
Medicaid and 
other payers 







General access of 
Medicaid paUents 
lo care in certain 
bospiUlt could 
decrease 


• Increasing the relatjve 
number o( privaU pay 
and otiier 'prodtable 
pa tjent groups 







Public hospitals 
could suder 
additional losses 



Stales could be 
(orced lo make 
additional ad- 
justments in 
Medicaid rates 
(or public 
hospitals 



Medicaid costs 
could Increase 



r 



hospitals. Whether this will increase the Medicaid caseload in 
relation to other, types of patients is not yet clear. If 
hospitals are particularly successful in increasing their numbers 
of non-Medicaid patients, it could result in reduced availability 
of hospital care to Medicaid patients, who would have to be 
acconmiodated elsewhere. 

(2) Increasing specialization of services — It is generally 
agreed that some hospital services are more profitable than 
others — for example, some basic services (obstetrics, for 
example) may be money losers, or "loss leaders", as several of 
our interviewees termed them. Depending on the ownership and 
basic mission policy of each institution, the administration will 
be more or less aggressively trying to increase income and 
minimize losses through development or limitation of selected 
services. The voluntary and proprietary hospitals will be freer 
to do this than the public hospitals, which can turn no one away 
and are forced to "carry" other facilities in their areas. 

Specialization could be beneficial, in that it would 
decrease unnecessary duplication of services and may lead to more 
efficient service delivery. However, it could also be 
undesirable, in that it could result in necessary services being 
unavailable or inaccessible, and it could be discriminatory as 
well. Certain population groups, such as women and blacks, have 
service needs that are different — and use some services more 
than other population groups. Eliminating these services can be 
discriminatory — i.e., if closing a service affects mostly a 
black population and prevents them from getting needed health 
care, they may be able to bring a class action suit on the basis 

73 



<_>' * 



of racial discrimination. 

Physician practices will be examined in the same light, and 
pressure will be brought on the "loss leaders" to change the 
medical practices perceived to be causing the losses. Again, the 
results may or may not be desirable or discriminatory. 



'f^- 



74 



ens LIBRflRV 




3 aD^S DDDD5a7t"^